In recent years, the manufacturing industry has been seeking advanced technologies to produce complex metallic components with high efficiency, low cost, and reliable performance. Among these technologies, 3D printing, also known as additive manufacturing, is regarded as a revolutionary approach that is insensitive to part complexity. When combined with traditional casting, 3D printing sand molds enable the rapid fabrication of complex sand casting parts, especially for aluminum alloy components used in automotive, aerospace, and marine applications. This dissertation focuses on the simulation and experimental investigation of the 3D printing sand mold casting process for complex parts, with a particular emphasis on parameter optimization, mold design, and the validation of an automotive engine cylinder block.
The main objective of this work was to study the influence of 3D printed sand mold process parameters on the microstructure and mechanical properties of cast aluminum alloy components. In addition, a conformal gating system for an engine cylinder block was designed and optimized through numerical simulation, and the resulting process was applied to produce actual cylinder block castings. This research demonstrates that 3D printing sand casting technology is a viable and advantageous route for manufacturing complex sand casting parts with short lead times and excellent quality.
1. Introduction and Background
Traditional manufacturing processes such as conventional sand casting are highly sensitive to the complexity of the component geometry. Complex castings often require multiple sand cores, intricate gating systems, and extensive tooling, which increases development time and cost. In contrast, 3D printing sand molds eliminates the need for patterns and core boxes, allowing the direct fabrication of molds and cores from digital models. This capability offers unprecedented design freedom and enables the production of integrated, lightweight, high-value sand casting parts. Moreover, the combination of 3D printing with conventional gravity casting can fulfill the requirements of low-volume production, rapid prototyping, and customized manufacturing.
The present study was carried out to address several key issues in the field of 3D printed sand casting:
- To quantitatively compare the casting quality achieved with 3D printed sand molds versus conventional no-bake sand molds and permanent metal molds.
- To evaluate how sand mold preheating, pouring temperature, and layer thickness during mold printing affect the microstructure and mechanical properties of the resulting aluminum castings.
- To design and simulate a conformal gating system for an automotive engine cylinder block, and to verify the optimal casting process using ProCAST simulation.
- To produce a defect-free engine cylinder block using 3D printed sand molds and to assess its performance against industry standards.
The term “sand casting parts” is used throughout this thesis to emphasize the target products of this technology. These parts are typically characterized by complex internal cavities, thin walls, and high requirements for dimensional accuracy and mechanical integrity.
2. Materials and Experimental Procedures
2.1 Alloy and Melt Preparation
The material selected for both the trial castings and the final cylinder block was ZL101A (AlSi7Mg0.3) aluminum alloy. The nominal chemical composition of ZL101A is given in Table 1.
| Si | Mg | Ti | Al |
|---|---|---|---|
| 6.5 – 7.5 | 0.25 – 0.45 | 0.08 – 0.20 | Balance |
In the actual melt, strontium (Sr) was added as a modifier to refine the eutectic silicon morphology and improve melt fluidity. The measured composition of the alloy used in the experiments is listed in Table 2.
| Si | Mg | Ti | Sr | Fe | Zn | V | Al |
|---|---|---|---|---|---|---|---|
| 7.223 | 0.322 | 0.106 | 0.041 | 0.109 | 0.017 | 0.015 | Balance |
The alloy was melted in an induction melting furnace with a capacity of 70 kg. Degassing was performed using argon gas with a rotary impeller, and slag was removed carefully. After modification with Sr, the melt was held for 30 minutes at 740 °C before pouring.
2.2 3D Printing of Sand Molds
The sand molds were produced using a PCM-800 sand mold 3D printer developed by our collaborating enterprise. This equipment employs binder jetting technology, in which furan resin droplets are selectively deposited onto a catalyst-mixed sand bed. The printing parameters used in this study are summarized in Table 3.
| Parameter | Value |
|---|---|
| Layer thickness | 0.4 mm or 0.5 mm |
| Printing speed | 25 s/layer |
| Printing mode | Single pass |
| Sand type | High-white silica sand |
| Binder | Furan resin, 2.36 wt.% |
| Catalyst | Sulfonic acid, 0.4 wt.% |
The properties of the printed sand specimens were tested according to standard procedures. The results are shown in Table 4. The printed sand molds exhibited adequate tensile, compressive, and bending strengths, as well as acceptable permeability and gas evolution values, making them suitable for gravity casting of aluminum alloys.
| Property | Value |
|---|---|
| Cold tensile strength | 0.8 – 1.5 MPa |
| Cold compressive strength | > 2 MPa |
| Cold bending strength | 1 – 2.5 MPa |
| Permeability | 80 – 120 |
| Gas evolution | 12 – 14 mL/g |
After printing, the molds were dried at 150 °C for 2 hours to remove residual moisture and volatile organic compounds. For some experiments, the molds were preheated to different temperatures before pouring. The standard casting process used a pouring temperature of 720 °C and a mold temperature of either room temperature or 100 – 200 °C.
3. Influence of Mold Type on Casting Quality
3.1 Experimental Design
To understand how the mold type affects the microstructure and mechanical properties of sand casting parts, three types of molds were prepared: conventional no-bake sand mold, 3D printed sand mold, and permanent metal mold. The pouring temperature was 720 °C, and the molds were not preheated. The casting had a wall thickness of 30 mm and was made from ZL101A. Test specimens were machined from identical locations in each casting for comparative analysis.
3.2 Microstructural Comparison
The as-cast microstructures of specimens from the three mold types are shown in Figure 1. The secondary dendrite arm spacing (SDAS) was measured using the intercept method, calculated by:
$$d=\frac{1}{m}\sum_{i=1}^{m}\frac{l_i}{n_i-1} \tag{1}$$
where \(d\) is the SDAS (μm), \(l_i\) is the length of the applied test line, and \(n_i\) is the number of dendrite arm intersections.
The measured SDAS values were 63.05 μm for the conventional sand mold, 84.08 μm for the 3D printed sand mold, and 25.94 μm for the metal mold. The larger SDAS in the 3D printed sand mold is attributed to the lower thermal conductivity of the sand material, which slows cooling compared with a metal mold. However, the 3D printed sand mold produced a more uniform microstructure with fewer porosity defects and inclusions, thanks to the homogeneous density and controllable properties of the printed mold.
3.3 Mechanical Properties
Tensile tests were performed on as-cast specimens. Table 5 compares the tensile strength and elongation for each mold type.
| Mold type | Ultimate tensile strength (MPa) | Elongation (%) |
|---|---|---|
| Conventional sand mold | 127 | 1.9 |
| 3D printed sand mold | 131 | 2.2 |
| Permanent metal mold | 184 | 3.4 |
Although the metal mold produced the highest strength and ductility, the 3D printed sand mold slightly outperformed the conventional sand mold in both strength and elongation. This improvement is due to better mold consistency and reduced gas-related defects. Fracture surface examination by SEM revealed a mixed ductile-brittle fracture mode in the 3D printed sand mold specimen, while the conventional sand mold specimen exhibited more cleavage facets, indicating a more brittle behavior.
3.4 Density Measurements
The density of the specimens was measured using the Archimedes method:
$$\rho_{\text{measured}} = \frac{m_A\,\rho_{\text{water}}}{m_A – m_B} \tag{2}$$
where \(m_A\) is the mass in air, \(m_B\) is the apparent mass when submerged in water, and \(\rho_{\text{water}}\) is the density of water. The results are shown in Table 6.
| Mold type | Density (g/cm³) |
|---|---|
| Conventional sand mold | 2.6260 |
| 3D printed sand mold | 2.6300 |
| Permanent metal mold | 2.6753 |
These results indicate that the 3D printed sand mold yields denser castings than conventional sand molds, which is beneficial for the integrity and pressure tightness of sand casting parts.
4. Effect of Process Parameters on 3D Printed Sand Mold Castings
4.1 Effect of Mold Preheating Temperature
To investigate the effect of mold preheating, 3D printed sand molds were heated to room temperature, 100 °C, 150 °C, and 200 °C before pouring. The pouring temperature was held constant at 720 °C. After casting, all samples received T6 heat treatment (solution at 530 °C for 8–10 h, quenched in warm water, and aged at 180 °C for 6–8 h).
Figure 2 shows the microstructures and the measured SDAS values. With increasing mold preheat temperature, the SDAS increased from 57.58 μm at room temperature to 90.72 μm at 200 °C. This coarsening is caused by slower cooling and longer solidification time at higher mold temperatures. The tensile test results after T6 are presented in Table 7.
| Mold preheat temperature | UTS (MPa) | Elongation (%) |
|---|---|---|
| Room temperature | 218 | 0.4 |
| 100 °C | 213 | 0.4 |
| 150 °C | 210 | 0.3 |
| 200 °C | 205 | 0.3 |
Therefore, for ordinary aluminum castings, preheating of the dried 3D printed sand mold is unnecessary. Pouring at room temperature results in finer microstructures and better mechanical properties.
4.2 Effect of Pouring Temperature
Using 3D printed sand molds at room temperature, aluminum melts were poured at 700 °C, 710 °C, 720 °C, and 740 °C. The T6-treated tensile properties are summarized in Table 8.
| Pouring temperature (°C) | UTS (MPa) | Elongation (%) |
|---|---|---|
| 700 | 210 | 0.3 |
| 710 | 220 | 0.4 |
| 720 | 229 | 0.5 |
| 740 | 221 | 0.4 |
The optimal pouring temperature is 720 °C. At this temperature, the melt has sufficient fluidity to fill thin sections, and the eutectic silicon is well modified into a fine, spherical morphology. At higher temperatures, iron-rich intermetallic phases such as β-Fe (Fe₂Si₂Al₉) and π-Fe (Al₈FeMg₃Si₆) are more likely to precipitate, which degrade the mechanical properties.
4.3 Effect of Printing Layer Thickness
Two sand molds were printed with layer thicknesses of 0.4 mm and 0.5 mm, respectively, while keeping all other printing parameters constant. The total casting height was 250 mm, resulting in 625 and 500 layers, respectively. A smaller layer thickness increases the amount of binder because more layers are required, leading to a higher gas evolution during pouring.
The T6-treated microstructures are compared in Figure 3. The 0.4 mm layer mold gave a smaller SDAS of 57.60 μm, while the 0.5 mm layer mold yielded a larger SDAS of 70.70 μm. However, the 0.4 mm layer mold produced more gas porosity because of the higher binder content. The density and tensile properties are listed in Table 9.
| Layer thickness | Density (g/cm³) | UTS (MPa) | Elongation (%) |
|---|---|---|---|
| 0.4 mm | 2.6454 | 218 | 0.4 |
| 0.5 mm | 2.6503 | 229 | 0.5 |
The 0.5 mm layer thickness is more favorable for producing sound sand casting parts, as it balances mold strength, permeability, and minimized gas defects. This layer thickness was therefore selected for the engine cylinder block production.
5. Design and Simulation of the Engine Cylinder Block Casting Process
5.1 Cylinder Block Description and Process Challenges
The target component is a three-cylinder inline automotive engine cylinder block. Its overall dimensions are approximately 500 × 500 × 500 mm, and its wall thickness varies from 2 mm to 40 mm. The thinnest region is the gap between adjacent cast-iron cylinder liners. The cylinder block houses three cast-iron liners with an outer diameter of 87.6 mm and a wall thickness of 6 mm. After casting, the liner outer diameter must not be less than 83 mm, and the liner must be fully bonded with the aluminum matrix. This component is a typical high-complexity sand casting part, requiring careful gating design and thermal control.
5.2 Conformal Gating System Design
Utilizing the capability of 3D printing, a conformal gating system was designed to optimize melt flow and directional solidification. The design was based on the “large orifice discharge” theory. The cross-sectional area ratio of the gating system was chosen as:
$$A_{\text{sprue}} : A_{\text{runner}} : A_{\text{ingate}} = 1 : (2 \sim 4) : (2 \sim 4) \tag{3}$$
The ingate area was calculated as:
$$A_{\text{ingate}} = \frac{G_L}{\rho_L \mu t \sqrt{2 g h_p}} \tag{4}$$
where \(G_L\) is the mass of melt passing through the lowest section, \(\rho_L\) is the melt density, \(\mu\) is the flow loss coefficient, \(t\) is the pouring time, \(g\) is the acceleration due to gravity, and \(h_p\) is the effective pressure head.
Three different casting orientations were compared, as described in Table 10.
| Plan | Orientation | Description | Total poured mass (kg) | Yield (%) |
|---|---|---|---|---|
| A | Upright | Bottom gating at thick bearing area | 46.99 | 38.52 |
| B | Sideways | Gating at the thickest portion | 40.86 | 44.30 |
| C | Inverted | Gating at the cylinder liner area | 48.17 | 37.58 |
In Plan C, the melt enters first at the bottom liner area, allowing direct bonding between the hot aluminum and the preheated liners. This orientation also facilitates a uniform upward filling pattern and ensures that the top runners and risers remain active for feeding during solidification.
5.3 Simulation Setup
The models were meshed with tetrahedral elements. The mesh size for the casting, gating system, and accessories was 4 mm, while the mold body was meshed with 20 mm elements. The total number of volume elements was 11,717,789. The heat transfer coefficients at the relevant interfaces are listed in Table 11.
| Interface | HTC (W/m²·°C) |
|---|---|
| Casting / sand mold | 500 |
| Casting / cast iron liner | 2000 |
| Cast iron liner / sand mold | 500 |
The initial temperatures were set as: pouring melt at 730 °C, liners preheated to 400 °C, mold at room temperature. The pouring rate was calculated from the total mass and filling time of 16 s, giving a mass flow rate of approximately 2.93 kg/s.
5.4 Filling and Solidification Results
The filling simulation showed that Plan A filled the main casting region in 11.20 s, Plan B in a comparable time but with more turbulence, and Plan C in 10.67 s. Plan C provided the smoothest filling front, with less oxidation and no obvious vortex. The solidification sequences are shown in Figure 4.
Plan A solidified the main body between 335 s and 440 s, with isolated liquid regions disappearing at 608 s (84.7% solid). Plan B showed large time differences in solidification between thick and thin sections, creating a risk of shrinkage defects. Plan C displayed the most uniform solidification, with isolated liquid regions eliminated at 680 s when the overall solid fraction was 82.8%. This uniform behavior helps avoid hot spots and favors sound casting parts.
5.5 Defect Prediction
Shrinkage porosity was predicted using the Niyama criterion:
$$N = \sqrt{\frac{G}{R}} \tag{5}$$
where \(G\) is the temperature gradient and \(R\) is the cooling rate. A low value of \(N\) indicates a high risk of microporosity. The simulation results for the three plans are compared in Table 12.
| Plan | Total defect volume (cc) | Average porosity (%) | Remarks |
|---|---|---|---|
| A | 0.913 | 12.503 | Minor defects in water jacket area |
| B | 1.043 | 12.620 | Large shrink at assembly hole |
| C | 0.899 | 11.102 | Defects outside critical zones |
Plan C was selected as the optimal process because it produced the smallest defect volume and the defects were located in nonfunctional regions. The final casting parameters are summarized in Table 13.
| Parameter | Value |
|---|---|
| Mold temperature | Room temperature |
| Liner preheat temperature | 400 °C |
| Pouring temperature | 730 °C |
| Pouring rate | 3 kg/s |
| Pouring time | 16 s |
| Sprue diameter | 60 mm |
| Static pressure head | 400 mm |
6. Rapid Production of the Engine Cylinder Block
6.1 Sand Mold Printing and Assembly
Based on the optimized plan, the sand mold for the engine cylinder block was designed as a single-piece assembly with integrated runners, risers, and venting channels. The mold was split into two halves for ease of cleaning and insertion of the cast-iron liners and chillers. The model was converted into 2D slice data with a layer thickness of 0.5 mm. The entire mold required 1215 layers and was printed in approximately 10 hours.
After printing, the mold was extracted from the build platform, cleaned with compressed air, and coated with a zirconia-based refractory coating. The mold was then dried at 150 °C for 2 hours. During assembly, chillers were placed at thick sections, and heating elements were installed to preheat the cylinder liners to 400 °C. The mold halves were joined and sealed to prevent leakage.
6.2 Casting and Post-Processing
The ZL101A alloy was melted in an industrial furnace, degassed, and modified with strontium. Gravity pouring was performed at 730 °C with a mass flow rate of 3 kg/s. After filling, the mold was left undisturbed for 10 minutes to allow complete solidification. The casting was then knocked out by gentle vibration. At this point, the production cycle from printing to demolding was only 3–4 days, which is dramatically shorter than conventional tooling-based approaches.
The casting was cut free from the gating system and subjected to T6 heat treatment: solution at 530 °C for 12 h, quenched in 70 °C water, and aged at 180 °C for 6 h. Finally, the part was machined to the required dimensions.
6.3 Quality Inspection
X-ray inspection of the cylinder block showed no significant internal defects. The cast-in liners were fully bonded to the aluminum matrix, with an average penetration depth of 2–3 mm. The outer diameter of the liners after casting was measured as 85.67 mm, which satisfies the requirement of being greater than 83 mm. Pressure testing at 0.3 MPa showed no leakage, confirming the absence of microporosity in critical areas.
6.4 Microstructure and Mechanical Properties
Metallographic samples were taken from the bearing cap area and the liner interface. The microstructure exhibited fine, uniformly distributed dendrites, with no visible segregation. The eutectic silicon was well refined after T6 treatment. The mechanical properties measured from separately cast test bars are compared with the standard requirements in Table 14.
| Property | Standard (ZL101A, T6) | Measured value |
|---|---|---|
| Ultimate tensile strength | > 220 MPa | 278 MPa |
| Elongation | > 2.0% | 2.3% |
| Hardness | 80 HB | 97 HB |
These results demonstrate that the 3D printed sand mold casting process is fully capable of producing high-quality sand casting parts with complex internal geometry, dimensional accuracy, and satisfactory mechanical performance.
7. Conclusions
This work systematically investigated the 3D printing sand mold casting process for complex aluminum components. The main conclusions are as follows:
- 3D printed sand molds produced castings with slightly improved tensile strength and elongation compared with conventional no-bake sand molds, because the uniform mold properties and reduced gas evolution led to fewer defects and higher density. The optimized printing layer thickness was 0.5 mm, which balanced mold strength and gas generation.
- Preheating the 3D printed sand mold before pouring coarsened the microstructure and reduced the mechanical properties. For ZL101A castings, pouring at 720 °C without mold preheating resulted in the best combination of strength (229 MPa after T6) and elongation (0.5%) for 30 mm wall thickness.
- A conformal gating system was designed for the engine cylinder block. Simulation results showed that the inverted orientation (Plan C), with the liners at the bottom and preheated to 400 °C, provided the most uniform filling and solidification. The predicted shrinkage porosity volume was only 0.899 cc, with defects confined to noncritical areas.
- Using the optimized 3D printed sand mold casting process, an engine cylinder block was successfully produced in only 3–4 days. The casting had no visible defects, excellent liner bonding, and passed pressure tightness tests. After T6 treatment, the tensile strength reached 278 MPa, elongation was 2.3%, and hardness was 97 HB, exceeding the standard requirements.
The research confirms that 3D printing sand molds offer a transformative approach for manufacturing complex sand casting parts, enabling rapid prototyping, low volume production, and design innovation without significant cost penalties. The process is particularly suitable for automotive and aerospace components that require high dimensional accuracy, complex internal passages, and reliable mechanical performance.