Our investigation focuses on the widely used AlSi10Mg alloy, examining the distinct pathways of Selective Laser Melting (SLM) additive manufacturing and binder jetting of sand molds (3DP) followed by traditional casting. This work stems from the growing application of these technologies in automotive sectors for prototyping, low-volume complex parts, and tooling. We aim to provide a detailed, first-person account of the microstructural evolution and resultant mechanical properties, placing particular emphasis on understanding the characteristics and limitations of sand casting parts produced via 3DP-printed molds.
The motivation is clear: while SLM promises high performance and design freedom, sand casting via 3DP molds offers a route for larger, more traditionally shaped components. A direct comparison of their as-processed and heat-treated states is crucial for informed material and process selection. We delve into the anisotropic behavior of SLM parts, the effectiveness of post-process heat treatments, and the inherent variability in sand casting parts.
1. Materials and Experimental Methodology
Our study employed gas-atomized AlSi10Mg powder for the SLM process. The chemical composition is standard, as summarized in Table 1. The powder morphology and properties were suitable for reliable layer-wise deposition.
| Si | Mg | Fe | Cu | Mn | Ni | Ti | Al |
|---|---|---|---|---|---|---|---|
| 10.20 | 0.31 | 0.06 | <0.01 | <0.01 | <0.01 | <0.01 | Bal. |
1.1 SLM Additive Manufacturing Process
We fabricated test bars using an SLM machine with a build volume of 425 mm × 425 mm × 20 mm. A critical aspect of our design was to investigate building orientation effects. Specimens were arranged at angles of 0°, 45°, and 90° relative to the build plate. The 45° orientation required support structures to ensure successful overhang fabrication, as illustrated in the build layout. The primary SLM parameters that govern melt pool dynamics and cooling rates are listed in Table 2.
| Parameter | Value |
|---|---|
| Layer Thickness | 40 µm |
| Laser Power | 400 W |
| Scan Speed | 1300 mm/s |
| Hatch Spacing | 150 µm |
| Base Plate Temperature | 160 °C |
A subset of the as-built SLM specimens underwent a stress-relief annealing treatment at 300°C for 2 hours, followed by furnace cooling. This treatment, represented by the thermal cycle $T(t) = 300°C \text{ for } t \geq 2h$, was intended to relieve residual stresses and modify the metastable microstructure.
1.2 Sand Casting via 3DP Molds
In parallel, we produced sand casting parts using molds fabricated by binder jetting (3DP). The mold material was silica sand (100-140 mesh) bonded with resin. The mold design included a gating system to facilitate smooth metal filling. The AlSi10Mg alloy was melted and poured into these 3DP sand molds to produce cast test bars of identical nominal dimensions to the SLM ones.
These sand casting parts were subjected to a full T6 heat treatment to attain peak strength. The solution treatment involved a stepped heating process to 535°C, holding for 10 hours, followed by water quenching. This was immediately followed by artificial aging, also using a stepped profile, with a peak temperature of 160°C for 8 hours. The purpose was to dissolve soluble phases and subsequently precipitate strengthening Mg₂Si particles, a process governed by diffusion kinetics: $D = D_0 \exp(-Q/RT)$, where $D$ is the diffusion coefficient, $Q$ the activation energy, and $T$ the absolute temperature.
1.3 Characterization Techniques
Metallographic samples were prepared from both SLM and sand casting parts for microstructural analysis using optical microscopy. Mechanical properties were evaluated via tensile testing according to standard methods. Hardness measurements and fractography using scanning electron microscopy (SEM) complemented the tensile data to understand failure mechanisms.
2. Microstructural Analysis: A Tale of Two Processes
The microstructures of the AlSi10Mg alloy produced by the two routes were fundamentally different, rooted in their vastly different solidification conditions.
2.1 SLM Microstructure (As-Built and Annealed)
The as-built SLM structure is characterized by extremely fine cellular-dendritic α-Al grains supersaturated with Si, surrounded by a nanoscale eutectic Si network. This results from the ultra-high cooling rates inherent to SLM, estimated to be in the range of $10^3$ to $10^6$ K/s. The melt pool boundaries are often visible.
After the 300°C/2h anneal, significant microstructural evolution occurs. The supersaturated α-Al matrix experiences Si diffusion, leading to the spheroidization and coarsening of the eutectic Si phase. The once-continuous network breaks into discrete, fine Si particles. The cellular solidification structure becomes less distinct. The annealing temperature, $T_{ann}$, is sufficient for diffusion-driven coarsening but below the alloy’s solidus, preventing liquidation. The driving force for spheroidization is the reduction of interfacial energy, $\Delta G = \gamma \Delta A$, where $\gamma$ is the interfacial energy and $\Delta A$ is the reduction in surface area.
2.2 Sand Casting Parts Microstructure (T6 Condition)
In stark contrast, the microstructure of the T6 heat-treated sand casting parts is coarse and distinctly different. The primary α-Al dendritic structure is clearly visible, with dendrite arm spacing (DAS) several orders of magnitude larger than the SLM cell size. This is a direct consequence of the slow cooling rate during solidification in the sand mold, typically between 0.1 and 10 K/s.
While the T6 treatment (solutionizing and aging) dissolves some of the as-cast eutectic Si and precipitates fine Mg₂Si, it cannot erase the coarse initial structure. The remaining eutectic Si phase is predominantly in a modified, yet often acicular (needle-like) or blocky morphology, distributed along the interdendritic regions. The presence of these coarse Si particles in sand casting parts is a critical factor influencing their mechanical behavior, especially ductility. The final microstructure can be summarized as: Coarse α-Al dendrites + (Modified Acicular/Blocky Eutectic Si) + Fine β”-Mg₂Si Precipitates.
The key microstructural differences are consolidated in Table 3.
| Feature | SLM (Annealed) | Sand Casting Parts (T6) |
|---|---|---|
| α-Al Matrix | Very fine, former cellular structure; homogeneous. | Coarse dendritic structure; clear chemical segregation. |
| Si Phase Morphology | Fine, discrete, spheroidized particles. | Coarse, acicular/blocky particles in interdendritic regions. |
| Primary Strengthening | Grain boundary & Si particle strengthening. | Precipitation hardening (Mg₂Si) + grain size. |
| Scale | Micron/sub-micron cellular/particle size. | Dendrite Arm Spacing (DAS) > 50 µm. |
3. Mechanical Performance and Anisotropy
The mechanical properties directly reflect the underlying microstructures, with SLM exhibiting superior but anisotropic behavior and sand casting parts showing lower, more isotropic, but variable performance.
3.1 SLM: Orientation Dependence and Annealing Effect
A striking finding was the significant anisotropy in the as-built SLM specimens. The tensile strength decreased systematically with increasing build angle. The relationship can be empirically described as:
$$ \sigma_{UTS}(\theta) \approx \sigma_{UTS}(0°) – k \cdot \theta $$
where $\theta$ is the build angle (0°, 45°, 90°), and $k$ is an empirical constant (~20 MPa/45° in our study).
The 0° (vertical) specimens exhibited the highest strength, followed by 45°, with 90° (horizontal) being the lowest. This anisotropy is attributed to two main factors: 1) The alignment of lack-of-fusion pores or micro-voids parallel to the build layers, creating weaker paths perpendicular to the build direction (90°), and 2) Better thermal conduction and interfacial bonding when depositing new layers onto a solid substrate below (0°) compared to depositing onto supports or overhangs (45°). The effective thermal conductivity $k_{eff}$ in the build direction differs from that perpendicular to it, influencing solidification.
The 300°C/2h annealing treatment effectively eliminated this mechanical anisotropy, homogenizing the strength across all orientations. However, this came at the cost of a reduction in yield strength ($\sigma_y$), particularly for the stronger orientations (0° and 45°), due to stress relief and microstructural coarsening. The change in yield strength due to annealing can be related to dislocation density reduction: $\Delta \sigma_y \propto G b \sqrt{\Delta\rho}$, where $G$ is the shear modulus, $b$ the Burgers vector, and $\rho$ the dislocation density.
Elongation ($\epsilon_f$), a measure of ductility, showed a remarkable and beneficial trend. Post-annealing, the ductility increased by a factor of 2.5 to 3 across all orientations. The 0° annealed specimens achieved elongations exceeding 10%, indicating good toughness. The data is summarized in Table 4.
| Build Orientation | As-Built | Annealed (300°C/2h) | ||||
|---|---|---|---|---|---|---|
| Yield Strength (MPa) | UTS (MPa) | Elongation (%) | Yield Strength (MPa) | UTS (MPa) | Elongation (%) | |
| 0° | ~295 | ~395 | ~3.5 | ~230 | ~350 | >10.0 |
| 45° | ~280 | ~375 | ~3.0 | ~238 | ~345 | ~8.0 |
| 90° | ~254 | ~340 | ~2.5 | ~240 | ~338 | ~6.5 |
3.2 Sand Casting Parts: Performance and Scatter
The sand casting parts, even after T6 heat treatment, demonstrated markedly lower mechanical properties compared to the SLM material. The yield strength and ultimate tensile strength were approximately 30-40% lower than the annealed SLM values. More critically, the elongation was severely limited, typically around 1.5%, which is only about 15-25% of the ductility seen in annealed SLM parts.
An important observation was the higher degree of scatter in the properties of the sand casting parts. This variability is intrinsic to the sand casting process and can be attributed to defects such as micro-porosity from gas entrapment, shrinkage cavities, and non-metallic inclusions. The probability of finding a critical defect of size $a$ in a volume $V$ influences the strength scatter according to Weibull statistics: $P_f = 1 – \exp[-(V/V_0)(\sigma/\sigma_0)^m]$, where $m$ is the Weibull modulus (lower $m$ indicates higher scatter). Sand casting parts inherently have a lower $m$ compared to wrought or finely processed materials like SLM.
The stepped heat treatment protocol was employed specifically to mitigate issues like mold gas evolution during rapid heating, which can exacerbate porosity in the final sand casting parts. The comparative data is presented in Table 5.
| Material & Condition | Yield Strength (MPa) | Ultimate Tensile Strength (MPa) | Elongation (%) | Property Stability |
|---|---|---|---|---|
| SLM (Annealed, 0°) | 230 | 350 | >10.0 | High |
| Sand Casting Parts (T6) | ~210 | ~240 | ~1.5 | Moderate to Low (Scatter) |
4. Fractography: Insights into Failure Mechanisms
Examination of fracture surfaces provided direct evidence for the differences in mechanical behavior.
SLM (Annealed): The fracture surface of the annealed SLM specimen (0° orientation) exhibited a mixed-mode appearance. Regions showed dimpled rupture, indicative of micro-void coalescence and ductile failure. Other areas displayed quasi-cleavage features with fine, faceted patterns and small tear ridges. The overall morphology suggests a moderately ductile fracture, consistent with the measured elongation >10%. Occasional fine pores were observed, acting as initiation sites.
Sand Casting Parts (T6): The fracture surface of the T6-treated sand casting parts was predominantly brittle. It was characterized by large, flat facets corresponding to cleavage of the coarse α-Al dendrites. The brittle, acicular eutectic Si particles were clearly visible, often acting as stress concentrators and crack initiation points. The path of the crack frequently followed the interdendritic regions where these coarse Si particles and other brittle phases are located. The lack of significant plastic deformation on the fracture surface correlates with the low elongation (~1.5%) measured in these sand casting parts.
5. Discussion: Process-Structure-Property Relationships
The disparity in performance between SLM-processed AlSi10Mg and sand casting parts can be fundamentally traced to the process-induced microstructure.
The SLM process, with its high energy density and rapid solidification, produces a fine, homogeneous, and metastable structure. The Hall-Petch relationship, $\sigma_y = \sigma_0 + k_y d^{-1/2}$, explains part of the high strength, where $d$ is the effective grain or cell size. The fine dispersion of Si particles provides additional strengthening via the Orowan mechanism: $\Delta \tau \approx (Gb/\lambda) \ln(d_p/b)$, where $\lambda$ is the interparticle spacing and $d_p$ the particle diameter. Annealing relieves stresses and improves ductility by spheroidizing the Si, reducing stress concentrations, though it slightly coarsens the structure and reduces yield strength.
For sand casting parts, the slow cooling results in a coarse dendritic structure with large DAS. The strength from precipitation hardening ($\beta”$-Mg₂Si) is partially offset by the coarse grain size (inverse Hall-Petch). The primary limitation, however, is ductility and toughness. The coarse, acicular Si particles are potent crack initiators. The fracture toughness $K_{IC}$ is severely compromised by such large, brittle secondary phases. Furthermore, the inherent process defects like porosity act as pre-existing cracks, further reducing the effective load-bearing area and promoting early failure. The performance of sand casting parts is thus intrinsically capped by these microstructural features, which even aggressive T6 heat treatment cannot fully rectify.
The anisotropy in as-built SLM parts is a direct consequence of the layer-wise additive process and the associated thermal gradients. Modeling the thermal history, governed by the heat conduction equation $ \rho C_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + Q_{laser} $, reveals different thermal histories for different orientations, leading to varied microstructures and properties.
6. Conclusion
This first-person comparative study conclusively demonstrates the profound influence of manufacturing route on the AlSi10Mg alloy. Selective Laser Melting produces components with superior and tailorable mechanical properties—high strength, excellent ductility after annealing, and high reproducibility. The initial anisotropy in as-built parts can be effectively mitigated through a simple stress-relief anneal, yielding isotropic and tough components suitable for demanding, lightweight structural applications.
Conversely, sand casting parts produced via 3DP-printed molds, even with optimized T6 heat treatment, exhibit significantly lower mechanical performance, particularly in terms of ductility and toughness. Their properties are limited by the coarse as-cast microstructure containing brittle Si phases and are subject to greater statistical scatter due to inherent casting defects. Therefore, sand casting parts via this route are more suitable for applications where complex geometry or larger size is required, but where the service loads are primarily static and high ductility/fatigue resistance are not critical.
The choice between these two advanced manufacturing paths for AlSi10Mg must be guided by a thorough understanding of these process-structure-property linkages, balancing the exceptional performance of SLM against the geometric and economic advantages of sand casting for producing certain sand casting parts.