Processing and Thermal Diffusivity Measurements of Compositionally Graded Al-12Si to Al₂O₃ Structures

Abstract

The directed energy deposition (DED)-based additive manufacturing (AM) was used to create compositionally graded pure Al-12Si to pure Al₂O₃ structures varying the powder feed rates during deposition. Thermal diffusivity of Al-12Si+Al₂O₃ structures was reduced by >60% compared to pure Al-12Si. With a pure Al₂O₃ ceramic layer on Al-12Si+Al₂O₃, our results confirm the feasibility of designing and manufacturing metal-ceramic composites via AM with tailored thermal properties.

Graphical Abstract

Introduction

Additive manufacturing (AM) of compositionally and functionally gradient materials (FGMs) can offer site-specific properties at user-defined locations [1–3]. Such structures are regularly used in nuclear, aerospace, and biomedical applications, to name a few [4–5]. A gradient interface is typically manufactured for AM processed FGMs to avoid a sharp mismatch in composition and properties [6–7]. Previous research has shown AM of various Al-alloys [8–9]. AM of Al-alloys with a reinforcement phase has also been investigated, and mechanical property improvements have been realized [10–13]. The addition of an FGM zone between Al₂O₃ and 316 stainless steel showed a 50% reduction in thermal stresses [14]. Addition of an FGM zone between Al₂O₃ and Ti6Al4V via directed energy deposition (DED)-based AM also reduced the residual thermal stresses in the structure and made the parts possible manufacturability [15]. Selective laser melting (SLM) of TiB₂ added Al-12Si improved microhardness from 76 ± 0.5 HV in hot-pressed condition to 142 ± 6.0 HV in the AM’ed condition. Laser processing increased the yield strength from 144 ± 8 MPa to 247 ± 4 MPa [16]. In this work, processing of Al-12Si alloy dynamically transitioning to pure Al₂O₃ ceramic in a single structure was explored, keeping the variation in thermal performance in mind. Due to a significant thermal property mismatch between the Al-12Si and Al₂O₃, an FGM zone was processed between the pure metal and the pure ceramic phases. Such structures can decrease thermal diffusivity because of the outer ceramic layer, which is vital in space and nuclear applications [3].

Experimental Methods

Al-12Si powder (S-20, Valimet, Inc., CA, particle size 45 to 150 μm) and Al₂O₃ powder (AL-604, Atlantic Equipment Engineers, NJ, particle size 45 to 150 μm) were utilized in our Laser Engineered Net Shaping (LENS™-750, Optomec Inc., NM) AM system. A 3mm thick SS 316 plate was used as a substrate, and the O₂ level was controlled below 20 ppm during processing. Al-12Si and Al₂O₃ powders were placed in two different powder feeders, powder feeder 1 and 2, to allow multiple material deposition during the build. Figure 1a demonstrates LENS™ processing of Al-12Si/Al₂O₃ FGM. A 10-layer pure Al-12Si section with a layer thickness of 0.3 mm was first deposited using a laser power of 200 W and a laser scan speed of 1200 mm/min. Only powder feeder 1 (Al-12Si) was opened to fabricate this section. A 5-layer Al-12Si+Al₂O₃ compositionally graded transition was fabricated on top of the pure Al-12Si. A laser power of 300 W and a laser scan speed of 1200 mm/min were utilized, where both powder feeders were opened, and each powder feeder was adjusted on the fly. The layer thickness of the Al-12Si+Al₂O₃ section was 0.25 mm. A 10-layer pure Al₂O₃ section was built after the Al-12Si+Al₂O₃ using 350 W laser power and 1200 mm/min laser scan. Only powder feed 2 (Al₂O₃) was opened. The layer thickness of fabricating the pure Al₂O₃ was 0.25 mm. All processing parameters are summarized in Table 1. The laser power needed to vary from the Al-12Si to Al₂O₃ due to different laser-materials interactions, heat capacity, and the two materials’ thermal conductivities. Such variations also caused a difference in layer thickness.

Figure 1:

(a) Structural design and LENS™ processing of Al-12Si/Al₂O₃ FGM sample. (b) Low magnification image of as-fabricated Al-12Si/Al₂O₃/Al-12Si FGM sample.

Table 1:

Processing parameters of each sample.

Sample	Laser Power (W)	Scan Speed (mm/min)	Powder Feed Rate (g/min)		Layer Thickness (mm)	Number of Layers
			All2Si	Al2O3
Pure Al-12Si	200	1200	9.5	0	0.3	10
Al-12Si + Al2O3	Al-12Si section: 200	1200	9.5	0	0.3	10
	Al-12Si+Al2O3 section: 300		9.5→0	0→8.2	0.25	10
Al-12Si/Al2O3 FGM	Al-12Si section: 200	1200	9.5	0	0.3	10
	Al-12Si+Al2O3 section: 300		9.5→0	0→8.2	0.25	5
	Al2O3 section: 350		0	8.2	0.25	10

Samples were polished and etched using an etchant (2 ml HF, 3 ml HCl, 5 ml HNO₃, and 190 ml DI water) for microstructural observations using an SEM (FESEM, FEI-SIRION, OR). XRD was completed on separate samples processed via LENS™ using a Siemens D-500 Kristalloflex Diffractometer (Siemens AG, Germany) from 25–70 2θ° with a Cu-Kα source. Vickers microhardness measurement was done (Phase II, Model 900–391, NJ) on top surfaces with 100 g (pure Al-12Si and Al-12Si+Al₂O₃ regions) and 400 g (pure Al₂O₃) loads for 15s dwell time. Thermal diffusivity was directly measured using a Netzsch LFA 447 NanoFlash^® (NETZSCH, CA system over 550–300°C in 50°C increments.

Results and Discussion

Microstructures of Al-12Si had regions with large silicon precipitate (bright regions) and a wide range of grain sizes, Figure 2a. Etched microstructures revealed melt pool boundary lines, which consisted of primarily columnar grains oriented toward the heat source. At the center, these grains were a mix of columnar and equiaxed. Similar behaviors have also been reported where α-Al dendritic structures were observed near melt pool boundaries, and equiaxed α-Al was observed in other regions [17]. Fibrous eutectic silicon phase in Al-rich matrix was observed similar to SLM of an aluminum-silicon alloy [18]. The insets of Figure 2a and b showed that silicon precipitates ranged from smooth and elongated to needle-like. Al-12Si microstructures in the Al-12Si+Al₂O₃ region, Figure 2b, revealed an equiaxed grain structure with silicon phases distributed around the Al₂O₃ particles. Al₂O₃ shows low thermal conductivity and high laser power absorption ability. Large residual heat was generated at Al₂O₃ rich regions, which reduced the solidification rate and formed equiaxed microstructures. Similar behavior was noted in the SLM of Al-12Si-TiB₂ composites due to the secondary ceramic phase [16]. Dendritic microstructures were also observed in the Al-12Si+Al₂O₃ regions, Figure 2c. The Al-12Si+Al₂O₃ section was fabricated with dynamically adjusted processing parameters, where the initial layers were Al-12Si dominated. Additionally, the high laser power was also applied, which leads to a higher cooling rate to favor dendritic microstructures. These dendrites were grown toward the heat source and heavily dominated by silicon. Al₂O₃ particles in the Al-12Si+Al₂O₃ region were partially melted and formed a particle reinforced composite.

Figure 2:

Microstructures in Al-12Si samples (dark regions-aluminum and lighter regions-silicon). (a) Al-12Si; (b) equiaxed grains and (c) dendritic growth in the Al-12Si+Al₂O₃ samples. (d) XRD spectra; (e) SEM image of the interface between Al-12Si+Al₂O₃ section and pure Al₂O₃ section.(f)SEM image of pure Al₂O₃ section.

Phase identification, Figure 2d, revealed phases of aluminum (PDF# 40–0787) and silicon (PDF# 27–1402) in Al-12Si, as well as corundum (α-Al₂O₃) phase (PDF# 10–0173) in the Al-12Si+Al₂O₃, which has also been reported previously in LENS™ processed alumina [15]. No new phase formation was detected. Figure 2e and 2f show the final transition from Al-12Si+Al₂O₃ to Al₂O₃ layer, where the Al-12Si layer is fully integrated with Al₂O₃ without any cracking. The hardness value at the pure Al-12Si region was 50.7 ± 19.6 HV_0.1 and was gradually increased to 72.9 ± 14.7 HV_0.1 in the Al-12Si+Al₂O₃ region (Figure 3a). The microhardness value at the Al-12Si+Al₂O₃ region was increased to 1118.2 ± 156.8 HV_0.4 as more Al₂O₃ was introduced. Very high hardness was observed in the pure Al₂O₃ section at 1958.7 ± 173.8 HV_0.4.

Processing multi-material structures are inherently tricky. Various building strategies were attempted as transitioning from Al-12Si to Al₂O₃ was not trivial. Mismatch in material properties can cause cracking within the structure and failure. With only a small amount of ceramic reinforcement, the temperature distribution was still primarily determined by the highly conductive Al-12Si. High heat dissipation occurred before local temperatures could fully melt the Al₂O₃ particles. However, when more ceramic was added into the melt, most of the applied thermal energy in regions surrounding the melt pool was retained, and residual temperatures remained high to allow the melting of the Al₂O₃ particles. This required a constantly evolving power input for AM processing.

Adding Al₂O₃ onto Al-12Si deposit was initially tried but without success. The mismatch between material properties such as coefficient of thermal expansion (CTE) was too significant and caused delamination as CTE for aluminum and Al₂O₃ are 23.6 μm/m-°C and 8.52 μm/m-°C, respectively [19]. Therefore, a compositional gradation was implemented to decrease this mismatch and improve final part quality. This gradation technique has been shown to improve processability in other systems and transition from Inconel 718 to copper [6] and stainless steel to vanadium carbide [7]. Figure 3b shows the region’s elemental dot map reinforced with Al₂O₃ that bridged the gap between pure Al-12Si and pure Al₂O₃.

Figure 3:

(a) Microhardness profile at the sample’s cross-section. (b) Elemental mapping of Al₂O₃ addition into Al-12Si showing the ceramic-reinforcement in the Al-12Si+Al₂O₃ region. (c) Thermal diffusivity trends for pure Al-12Si, Al-12Si + Al₂O₃ and Al-12Si/Al₂O₃ FGM samples.

In the Al-12Si, thermal diffusivity ranged from 45–43 mm²/s over 50–300°C, Figure 3c. With the addition of Al₂O₃, the Al-12Si+Al₂O₃ thermal diffusivity decreased to 22–19 mm²/s. The addition of a pure Al₂O₃ on the Al-12Si+Al₂O₃ decreased thermal diffusivity even further to 21–15 mm²/s over 50–300°C, a >60% drop due to the addition of the ceramic phase. With the addition of an insulative reinforcement phase, it is clear that the thermal performance should decrease. This was observed by adding 40 vol% SiC reinforcement into pure aluminum, where thermal conductivity decreased by up to 40% [20]. From Figure 3c, it was also observed that Al-12Si and Al-12Si+Al₂O₃ followed a linearly decreasing trend while the Al₂O₃-coated Al-12Si+Al₂O₃ showed a continuously decreasing thermal diffusivity over the same range.

Conclusions

Additive manufacturing was used to create functionally graded composite structures of Al-12Si to Al₂O₃-reinforced Al-12Si to Al₂O₃. Microhardness increased from 50.7 ± 19.6 HV_0.1 in Al-12Si alloy to 1958.7 ± 173.8 HV_0.4 in Al₂O₃ ceramic. A complex microstructural variation across the Al-12Si and Al-12Si+Al₂O₃ regions resulted in a >60% reduction in thermal diffusivity due to Al₂O₃ addition in the 50–300°C temperature range. Future studies will involve studying the mechanical properties of these structures under different loading conditions.

Highlights.

• Processing of Al-12Si to Al₂O₃ ceramic structures using additive manufacturing.

• An intermediate Al12Si + Al₂O₃ FGM was created to minimize thermal stresses.

• Microhardness increased from 50.7 ± 19.6 (Al-12Si) to 1958.7 ± 173.8 HV_0.4 (Al₂O₃).

• A reduction of > 50% in thermal diffusivity due to Al₂O₃ addition to Al-12Si alloy.