In modern manufacturing, sand casting remains a cornerstone for producing complex metal parts due to its versatility and cost-effectiveness. The advent of additive manufacturing, particularly inkjet 3D printing (3DP), has revolutionized the fabrication of sand casting molds and cores, enabling rapid prototyping and production of intricate geometries without traditional pattern-making. This study focuses on optimizing process parameters for inkjet 3DP of sand casting molds and cores, with an emphasis on reusing reclaimed sand to reduce costs and environmental impact. Through orthogonal experiments, we investigate the effects of material composition and printing parameters on tensile strength and loss on ignition, aiming to establish guidelines for industrial applications in sand casting.
The integration of 3D printing into sand casting processes addresses limitations of conventional methods, such as long lead times and high costs for complex designs. Inkjet D printing, in particular, offers advantages over selective laser sintering (SLS) by eliminating thermal stresses and reducing equipment expenses. However, the mechanical properties and quality of printed molds and cores are highly dependent on parameters like reclaimed sand ratio, layer thickness, and binder distribution. Reusing reclaimed sand—a byproduct of post-printing cleanup—introduces challenges related to weak interfacial layers and variable binder adhesion, necessitating systematic study. Our research leverages orthogonal design to analyze these factors, providing insights into strength behavior and thermal decomposition for sand casting applications.
Sand casting relies on durable molds and cores to shape molten metal, and 3DP enables precise control over their internal structures. The binder jetting process involves depositing furan resin binder onto a powder bed of silica sand mixed with a catalyst, with layer-by-layer solidification. Key parameters include the proportion of reclaimed sand, which affects interfacial bonding; layer thickness, influencing interlayer cohesion and print efficiency; and X-direction resolution, governing binder droplet spacing and quantity. Understanding their interactions is crucial for achieving optimal strength and minimizing gas evolution during casting, a common issue in sand casting related to high loss on ignition. This paper details our experimental approach, results, and theoretical analysis to advance 3DP for sand casting.
Background and Significance
Sand casting is a ubiquitous method in foundries for producing components ranging from engine blocks to turbine blades. Traditional mold-making involves compacting sand around patterns, which can be time-consuming for complex designs. Additive manufacturing, especially binder jetting 3DP, allows direct digital fabrication of molds and cores, reducing steps and enabling customization. The process uses silica sand as the base material, with binder droplets selectively printed to bond particles. Reclaimed sand, collected after printing, contains residual catalyst and binder, altering its properties. Reusing it promotes sustainability but requires parameter adjustments to maintain performance in sand casting.
Previous studies have highlighted the importance of process parameters in 3DP for sand casting. For instance, layer thickness impacts surface finish and strength, while binder content affects dimensional accuracy and gas emissions. However, comprehensive analyses integrating reclaimed sand usage are limited. Our work fills this gap by employing orthogonal experiments—a statistical method efficient for multi-factor testing—to evaluate tensile strength and loss on ignition. These metrics are critical for sand casting: adequate strength ensures mold integrity during handling and pouring, while low loss on ignition reduces casting defects like porosity from gas evolution.
Materials and Experimental Methods
We utilized silica sand with a particle size of 70–140 mesh, typical for sand casting applications to ensure proper permeability and surface detail. The sand particles are irregularly shaped, enhancing interlocking and binder adhesion. Reclaimed sand was obtained from unused powder after printing, with a loss on ignition of approximately 0.28%, indicating residual organics. The binder was furan resin, and the catalyst was p-toluenesulfonic acid, common in sand casting for their fast curing and high strength. Equipment included an ExOne MAX inkjet 3D printer with a build volume of 1800 mm × 1000 mm × 700 mm, and a hydraulic tensile strength tester for mechanical evaluation.
The tensile strength specimens were printed according to a standard geometry, with dimensions as shown in the following representation: the specimen has a central cylindrical section for testing, ensuring uniform stress distribution. After printing, samples were cleaned and cured for 24 hours before testing. Loss on ignition was measured by burning samples at high temperature to combust organics, calculated as the mass loss percentage. We designed a three-factor, four-level orthogonal experiment to minimize trial numbers while capturing parameter effects. The factors and levels are summarized in Table 1.
| Level | (A) Reclaimed Sand Ratio (%) | (B) Layer Thickness (mm) | (C) X-Direction Resolution (mm) |
|---|---|---|---|
| 1 | 0 | 0.24 | 0.06 |
| 2 | 30 | 0.28 | 0.07 |
| 3 | 60 | 0.32 | 0.08 |
| 4 | 90 | 0.36 | 0.09 |
The orthogonal array L16 was used, resulting in 16 experimental runs. Each run involved printing tensile specimens and measuring strength and loss on ignition. The reclaimed sand ratio was varied from 0% to 90% by mass, representing full reuse scenarios. Layer thickness ranged from 0.24 mm to 0.36 mm, balancing resolution and speed. X-direction resolution, defined as the distance between binder droplet ejections along the print head path, controlled binder content; finer resolution increases binder amount per unit area. This parameter is vital in sand casting for ensuring adequate bonding without excess gas generation.
Theoretical Framework
The strength of 3D printed sand casting molds and cores stems from binder bridges formed between sand particles. When binder droplets impact the powder bed, they spread and penetrate due to capillary forces, described by the Washburn equation for porous media: $$ L = \sqrt{\frac{\gamma \cos \theta}{2 \eta} t} $$ where \( L \) is penetration depth, \( \gamma \) is surface tension, \( \theta \) is contact angle, \( \eta \) is viscosity, and \( t \) is time. This influences interlayer bonding and overall integrity.
For reclaimed sand, weak interfacial layers may form due to contaminants or residual catalyst, reducing adhesion. The tensile strength \( \sigma_t \) can be modeled based on binder bridge area and particle packing: $$ \sigma_t = k \cdot \frac{A_b}{V_p} \cdot \sigma_b $$ where \( k \) is a constant, \( A_b \) is total binder bridge area, \( V_p \) is particle volume, and \( \sigma_b \) is binder strength. Parameters like layer thickness and X-direction resolution affect \( A_b \), as thicker layers reduce bridge density and coarser resolution decreases binder coverage.
Loss on ignition \( LOI \) relates to binder content \( C_b \) and decomposition: $$ LOI = \alpha \cdot C_b + \beta $$ with \( \alpha \) and \( \beta \) as coefficients. In sand casting, high \( LOI \) leads to gas defects, so optimizing parameters to minimize it is crucial. The Laplace equation explains capillary pressure during binder penetration: $$ \Delta P = \frac{2 \gamma}{r} $$ where \( r \) is pore radius. As layer thickness increases, porosity rises, reducing \( r \) and \( \Delta P \), hindering penetration and weakening bonds—a key consideration for sand casting quality.
Results and Analysis
The orthogonal experiment results for tensile strength and loss on ignition are presented in Table 2. Each combination yielded distinct values, with the highest strength of 3.61 MPa for 0% reclaimed sand, 0.24 mm layer thickness, and 0.06 mm X-direction resolution. The lowest loss on ignition was 1.93% for 0% reclaimed sand, 0.36 mm layer thickness, and 0.09 mm resolution. We analyzed the data using analysis of variance (ANOVA) to determine factor significance.
| Run No. | A: Reclaimed Sand (%) | B: Layer Thickness (mm) | C: X-Direction Resolution (mm) | Tensile Strength (MPa) | Loss on Ignition (%) |
|---|---|---|---|---|---|
| 1 | 0 | 0.24 | 0.06 | 3.61 | 3.67 |
| 2 | 0 | 0.28 | 0.07 | 3.10 | 2.67 |
| 3 | 0 | 0.32 | 0.08 | 2.32 | 2.22 |
| 4 | 0 | 0.36 | 0.09 | 1.48 | 1.93 |
| 5 | 30 | 0.28 | 0.06 | 3.02 | 2.97 |
| 6 | 30 | 0.24 | 0.07 | 3.22 | 2.74 |
| 7 | 30 | 0.36 | 0.08 | 1.70 | 1.99 |
| 8 | 30 | 0.32 | 0.09 | 1.82 | 2.04 |
| 9 | 60 | 0.32 | 0.06 | 2.47 | 3.09 |
| 10 | 60 | 0.36 | 0.07 | 1.73 | 2.23 |
| 11 | 60 | 0.24 | 0.08 | 3.03 | 2.72 |
| 12 | 60 | 0.28 | 0.09 | 2.19 | 2.18 |
| 13 | 90 | 0.36 | 0.06 | 2.04 | 2.70 |
| 14 | 90 | 0.32 | 0.07 | 2.27 | 2.17 |
| 15 | 90 | 0.28 | 0.08 | 2.41 | 2.38 |
| 16 | 90 | 0.24 | 0.09 | 2.69 | 2.34 |
ANOVA for tensile strength, shown in Table 3, revealed that all factors are significant, with layer thickness having the largest F-value (217.18), followed by X-direction resolution (59.81) and reclaimed sand ratio (9.67). This indicates that layer thickness is the most influential parameter for strength in sand casting 3DP. For loss on ignition, Table 4 shows X-direction resolution as the most significant factor (F=64.82), then layer thickness (F=27.9), while reclaimed sand ratio is less impactful (F=3.89). These findings guide parameter optimization for sand casting applications.
| Source | Sum of Squares | Degrees of Freedom | Mean Square | F-value | Critical F (α=0.05) |
|---|---|---|---|---|---|
| Reclaimed Sand Ratio | 0.19371 | 3 | 0.06457 | 9.67 | 4.76 |
| Layer Thickness | 4.34845 | 3 | 1.44948 | 217.18 | |
| X-Direction Resolution | 1.19759 | 3 | 0.39920 | 59.81 | |
| Error | 0.04004 | 6 | 0.00667 | – | – |
| Source | Sum of Squares | Degrees of Freedom | Mean Square | F-value | Critical F (α=0.05) |
|---|---|---|---|---|---|
| Reclaimed Sand Ratio | 0.13103 | 3 | 0.04368 | 3.89 | 4.76 |
| Layer Thickness | 0.93854 | 3 | 0.31285 | 27.9 | |
| X-Direction Resolution | 2.18077 | 3 | 0.72692 | 64.82 | |
| Error | 0.06728 | 6 | 0.01121 | – | – |
Main effect plots were generated to visualize parameter impacts. Tensile strength decreases with increasing reclaimed sand ratio, layer thickness, and X-direction resolution. For example, strength drops from about 3.6 MPa at 0% reclaimed sand to 2.5 MPa at 90%, due to weak interfacial layers. Similarly, as layer thickness rises from 0.24 mm to 0.36 mm, strength declines by approximately 40%, attributed to reduced interlayer bonding. X-direction resolution shows a milder effect, with strength decreasing by 30% from 0.06 mm to 0.09 mm, linked to lower binder content. These trends are critical for designing sand casting molds and cores with adequate strength.
Loss on ignition decreases with higher layer thickness and X-direction resolution, but increases with finer resolution due to more binder. Reclaimed sand ratio has a minimal effect, as its organic content is low. This aligns with sand casting requirements, where low loss on ignition (below 2.2%) is preferred to minimize gas defects during metal pouring. Based on industrial standards for sand casting—tensile strength ≥1.75 MPa and loss on ignition ≤2.2%—we identified optimal parameter combinations. Runs 8, 12, and 14 meet these criteria, with run 12 (30% reclaimed sand, 0.28 mm layer thickness, 0.09 mm X-direction resolution) offering a balance of strength and low gas evolution.
Discussion on Parameter Effects in Sand Casting
The influence of reclaimed sand ratio on tensile strength is explained by interfacial adhesion. Reclaimed sand particles may have residual catalyst or contaminants, creating weak boundaries that promote adhesive failure over cohesive binder rupture. This reduces effective bonding area, modeled as: $$ A_{eff} = A_0 (1 – \phi) $$ where \( A_0 \) is ideal area and \( \phi \) is weakness factor proportional to reclaimed sand content. For sand casting, using up to 30% reclaimed sand maintains strength within acceptable limits, supporting sustainable practices without compromising mold integrity.
X-direction resolution directly controls binder droplet density. Finer resolution increases binder volume per layer, enhancing bridge formation but raising loss on ignition. The relationship can be expressed as: $$ C_b = \frac{V_d}{d_x \cdot d_y \cdot h} $$ where \( V_d \) is droplet volume, \( d_x \) is X-direction resolution, \( d_y \) is Y-direction resolution (constant), and \( h \) is layer thickness. As \( d_x \) decreases, \( C_b \) increases, boosting strength but also gas generation—a trade-off in sand casting. Our results show that 0.09 mm resolution provides sufficient strength while keeping loss on ignition low, ideal for sand casting applications.
Layer thickness affects particle packing and capillary penetration. Thicker layers increase porosity, reducing capillary pressure \( \Delta P \) as per Laplace equation, which hinders binder spread. This weakens interlayer bonds, decreasing strength. The effect on loss on ignition is inverse: thicker layers have less binder per unit volume, lowering organic content. For sand casting, a layer thickness of 0.28 mm optimizes both strength and print efficiency. The dominance of layer thickness in ANOVA underscores its critical role in 3DP process design for sand casting.
Microstructural analysis of binder bridges reveals two failure modes: cohesive fracture within the bridge and adhesive failure at particle interfaces. Reclaimed sand increases adhesive failures, reducing strength. We propose a strength model incorporating these modes: $$ \sigma_t = \sigma_c \cdot f_c + \sigma_a \cdot f_a $$ where \( \sigma_c \) and \( \sigma_a \) are cohesive and adhesive strengths, and \( f_c \), \( f_a \) their frequency. Parameters like layer thickness and resolution influence \( f_c \) by altering bridge geometry, vital for sand casting mold durability.
Practical Implications for Sand Casting Industry
Our findings provide actionable guidelines for implementing inkjet 3DP in sand casting foundries. By adopting a reclaimed sand ratio of 30%, layer thickness of 0.28 mm, and X-direction resolution of 0.09 mm, manufacturers can produce molds and cores with tensile strength around 1.97 MPa and loss on ignition of 2.2%, meeting typical sand casting specifications. This reduces material costs by reusing sand and minimizes gas-related defects, improving casting yield. The orthogonal approach efficiently identifies robust parameters, saving time in trial-and-error adjustments.
For complex sand casting components like engine blocks or impellers, these parameters ensure dimensional accuracy and strength. The 3DP process enables intricate cores impossible with traditional methods, enhancing design freedom. Additionally, controlling loss on ignition is crucial for high-quality sand casting, as excess gases can cause porosity in cast metals. Our study highlights the interplay between parameters, allowing foundries to tailor settings for specific alloys or geometries in sand casting.
Future Research Directions
While this study focuses on tensile strength and loss on ignition, other properties relevant to sand casting warrant investigation, such as permeability, thermal stability, and surface roughness. Future work could explore multi-objective optimization using response surface methodology or machine learning to balance multiple performance metrics. The effect of sand particle shape and size distribution on 3DP outcomes for sand casting should be examined, as irregular grains may enhance bonding but affect flowability. Long-term durability of printed molds under repeated thermal cycles in sand casting is another area for exploration.
Furthermore, environmental aspects like binder formulation and recycling processes can be optimized to reduce emissions in sand casting. Integrating real-time monitoring during printing could improve consistency. As sand casting evolves with additive manufacturing, our research lays a foundation for parameter standardization, promoting wider adoption of 3DP in foundries.
Conclusion
In this study, we investigated the forming process parameters for inkjet 3D printed sand casting molds and cores through orthogonal experiments. The results demonstrate that layer thickness has the most significant impact on tensile strength, followed by X-direction resolution and reclaimed sand ratio. For loss on ignition, X-direction resolution is the dominant factor. Increasing reclaimed sand ratio reduces strength due to weak interfacial layers, while thicker layers decrease both strength and loss on ignition. Finer X-direction resolution increases strength but also loss on ignition. The optimal parameters for sand casting applications are 30% reclaimed sand, 0.28 mm layer thickness, and 0.09 mm X-direction resolution, yielding a tensile strength of 1.97 MPa and loss on ignition of 2.2%, which satisfy industrial sand casting requirements. This research advances the understanding of 3DP for sand casting, enabling cost-effective and sustainable production of high-quality molds and cores.