In the pursuit of enhancing product performance and competitiveness, our company embarked on a project to optimize and upgrade an existing integral cylinder head casting. This specific casting part is a six-cylinder-in-one-head design, originally produced with mature and stable conventional sand casting processes using HT300 gray iron. The upgrade required significant modifications to the internal water jacket and intake/exhaust port geometries, coupled with a material change to the stronger RuT450 (compacted graphite iron). Adopting the traditional development model for such a casting part would necessitate the complete redesign, fabrication, and debugging of new core boxes and tooling. This approach incurs high trial production costs and leads to prohibitively long validation cycles, delaying time-to-market.
To overcome these challenges and supply test samples promptly, we leveraged sand core 3D printing technology for rapid prototyping. This method allowed us to directly fabricate complex sand cores based on the updated three-dimensional digital models, bypassing the need for physical tooling. The process flow involved printing the individual sand cores, assembling them, applying coating, and finally pouring the molten metal to produce the prototype casting parts.
Fundamental Information of the Casting Part
Casting Part Structure
The integral cylinder head casting part has overall dimensions of 768 mm × 220 mm × 118 mm, with a final casting weight of 84.8 kg. Its structure features thin walls, with a primary wall thickness of approximately 5 mm. The complexity lies in its internal cavities, which form the coolant passages (water jacket) and the intricate intake and exhaust ports for all six cylinders.
Existing Foundry Process Layout
The conventional production process for this cylinder head involves several sand cores: a tray core, lower water jacket core, intake port cores, exhaust port cores, upper water jacket core, and tappet bore cores. Each core is produced individually in its respective core box. Core assembly is performed using sodium silicate-based adhesive. The casting’s lower plane and all internal surfaces are formed by these assembled sand cores, while the upper surface is shaped by the cope half of the mold (external pattern). The molding is configured for two casting parts per mold box, arranged symmetrically. A bottom-gating system is employed, with ingates located in the tray core, allowing the molten iron to fill the mold cavity from the bottom upward.
Development and Trial Production of the Optimized Casting Part
Since the design optimization primarily altered the internal geometry of the casting part, the existing core boxes for the affected cores became obsolete. Utilizing 3D printed sand cores was the strategic choice for initial sample production. The external shape of the casting part remained unchanged, allowing us to continue using the existing external patterns for molding.
To mitigate risk and evaluate different material systems, we sourced 3D printed sand cores from two distinct suppliers, referred to here as Manufacturer A and Manufacturer B. Prior to full-scale trial production, we conducted a comprehensive comparative analysis of test specimens provided by both manufacturers. The evaluation covered fundamental physical properties, tensile strength, gas evolution characteristics, and high-temperature performance. All cores for trial production were coated with a water-based refractory coating (viscosity: 34 °Bé) and dried at 125°C for 2.5 hours.
Characterization of 3D Printed Sand Specimens
A rigorous testing regime was established to quantify the performance of the 3D printing sand materials, as their behavior directly dictates the quality and feasibility of producing sound casting parts.
Basic Physical Properties
The “dog-bone” tensile test specimens from both manufacturers were analyzed. Key observations are summarized in the table below. The standard core mixture used in production yielded a specimen mass of 66.3 g, heavier than both 3D printed versions. Manufacturer A’s sand blend had a lower average grain fineness number (AFS GFN) compared to Manufacturer B’s.
| Supplier | Mass (g) | Appearance | Primary Composition | Avg. Fineness (AFS) |
|---|---|---|---|---|
| Manufacturer A | 61.3 | Light Green | Silica Sand + Ceramic Beads | 91.2 |
| Manufacturer B | 63.9 | Dark Green | Ceramic Beads | 99.8 |
Tensile Strength Analysis
Tensile strength was measured at three stages: as-printed (“green”), after drying without coating, and after drying with coating applied. The drying cycle was 125°C for 2.5 hours. Adequate tensile strength is critical to prevent core fracture during mold assembly, handling, and the initial stages of metal pouring.
| Supplier | As-Printed (MPa) | Dried, No Coat (MPa) | Dried, With Coat (MPa) |
|---|---|---|---|
| Manufacturer A | 2.11 | 2.30 | 1.97 |
| Manufacturer B | 3.15 | 4.00 | 2.93 |
The data shows that drying increases strength, likely due to further curing of the binder. However, the coating process can slightly reduce the measured strength, possibly due to moisture introduction or a slight abrasive effect during dipping. Manufacturer B’s material exhibited consistently higher tensile strength, which is a favorable characteristic for maintaining core integrity during the production of complex casting parts.
Gas Evolution Behavior
The gas evolution volume and rate are paramount for casting part quality. Excessive or rapid gas generation can lead to defects like blows and pinholes if the mold and core cannot vent the gas quickly enough. We measured the total gas evolved and the evolution kinetics over 300 seconds at a standard temperature.
| Supplier | Dried, No Coat (mL/g) | Dried, With Coat (mL/g) |
|---|---|---|
| Manufacturer A | 15.49 | 16.55 |
| Manufacturer B | 21.20 | 23.72 |
Manufacturer B’s material demonstrated significantly higher gas evolution. The coating adds a small increment to the total gas volume. The gas evolution rate curves revealed that approximately 65% of the total gas was released within the first 50 seconds, and about 86% within 100 seconds, for both materials. This rapid early gas evolution imposes stringent demands on the venting capacity of the mold system during the critical initial filling stage for these casting parts.
The gas evolution process can be conceptually modeled. The total gas volume $V_{total}$ is a function of the binder and coating composition. The instantaneous gas evolution rate $dV/dt$ often follows a decay pattern after a sharp initial peak, which can be approximated by an exponential decay function related to the breakdown kinetics of the organic components:
$$ \frac{dV}{dt} = G_0 \cdot e^{-k t} $$
where $G_0$ is the initial peak gas evolution rate and $k$ is a rate constant dependent on the material and temperature.
High-Temperature Performance
Cores must withstand the thermal shock and sustained heat of molten metal without excessive expansion or loss of structural integrity (thermal collapse). Cylindrical specimens (φ12 mm × 20 mm) were heated to 1000°C to evaluate their thermo-mechanical behavior.
| Supplier | High-Temp Expansion (%) | Time to Peak Expansion (s) | High-Temp Durability (s) |
|---|---|---|---|
| Manufacturer A | 0.095 | 17.8 | 93.2 |
| Manufacturer B | 0.066 | 21.4 | 107.4 |
Manufacturer A’s material showed higher thermal expansion, reached its peak expansion faster, and had a shorter high-temperature durability time. This indicates a potentially lower resistance to thermal stress and deformation. Core expansion can lead to veining or hot tearing defects in the final casting parts, while early collapse can cause metal penetration or shape distortion.
The linear thermal expansion strain $\epsilon_{th}$ can be expressed as:
$$ \epsilon_{th} = \alpha \cdot \Delta T $$
where $\alpha$ is the coefficient of thermal expansion and $\Delta T$ is the temperature change. A lower $\alpha$ is generally desirable for dimensional stability of the core and, consequently, the casting part.
Trial Production of Casting Parts and Results
Initial Trial Production
The first trial run utilized 3D printed cores from both Manufacturers A and B. The process steps were followed: careful manual core assembly with enhanced adhesive application, dipping of the complete core assembly, drying, molding using the original external patterns, and pouring with RuT450 iron at 1410°C.
The results were unsatisfactory. Severe metal boiling (spitting) occurred during pouring. Upon shakeout, all casting parts exhibited severe core-related defects: the lower water jacket cores had shifted and fractured (“floated”), and the upper water jacket cores had broken at the newly designed fuel injector bore locations.
Root Cause Analysis and Process Optimization
Based on the initial failure, a multi-faceted root cause analysis was conducted, leading to several targeted countermeasures for subsequent trials:
- Gating Modification: To reduce direct molten metal impingement and washing on the vulnerable lower water jacket core, the ingate positions on the tray core were moved 8 mm inwards, away from the core walls.
- Core Design Integration: To improve mechanical connection and rigidity, the tray core and lower water jacket core were redesigned as a single, integrated 3D printed component.
- Use of Core Supports (Chills): Given the relatively thin sections and high buoyancy forces on the water jacket cores, core supports (chaplets) were strategically placed between the upper and lower water jacket cores in some trial schemes to counteract lifting forces. The buoyant force $F_b$ acting on a core is given by Archimedes’ principle:
$$ F_b = \rho_{metal} \cdot V_{submerged} \cdot g – \rho_{core} \cdot V_{core} \cdot g $$
where $\rho$ denotes density, $V$ volume, and $g$ gravity. This force must be resisted by the core’s strength and the mold/core interface. - Core Venting Enhancement: The new design included fuel injector bores in the upper water jacket core, which partially blocked the existing vent channels from the top of the mold. The geometry around these bores was optimized to enlarge the gas escape pathways, reducing pressure buildup on the core during pouring.
Structured Follow-up Trials and Results
A structured experiment was designed to systematically evaluate the improvements and compare the two core materials. Each mold box contained one set of original production cores (as a control) and one set of 3D printed cores, clearly marked for identification.
| Trial Scheme | Core Supplier | Core Supports Used? | Assembly Coated? | Assembly Dried? | Casting Part Quality Observations |
|---|---|---|---|---|---|
| 1 | A | Yes | Yes | Yes | Lower core floated; Upper core fractured at injector bore. |
| 2 | A | Yes | No | Yes | Lower core floated; No upper core fracture. |
| 3 | A | No | No | No | Lower core floated; No upper core fracture. |
| 4 | B | Yes | Yes | Yes | No floating; No fracture. Internal cavity sound. |
| 5 | B | No | No | Yes | No floating; No fracture. Internal cavity sound. |
| 6 | B | No | No | No | No floating; No fracture. Internal cavity sound. |
| 7 (Control) | Production Cores | No | Yes | Yes | No floating; No fracture. Internal cavity sound. |
The results were decisive. Casting parts produced with Manufacturer A’s 3D printed cores consistently suffered from lower water jacket core floating, regardless of the use of supports or coating. Furthermore, the coated and dried assemblies from Manufacturer A showed fracturing at the injector bore. In stark contrast, all casting parts produced with Manufacturer B’s 3D printed cores were free from these defects, even in the most minimal process condition (no supports, no coating, no post-printing drying). The internal surfaces of these casting parts were smooth, without burn-in or sand sintering, and the overall dimensional accuracy met specifications.
Subsequent verification trials using Manufacturer B’s cores under the “no coating, no drying” protocol confirmed the robustness of the process. Samples taken from the casting parts confirmed that the mechanical properties and microstructure of the RuT450 material met all requirements.
Conclusions and Implications for Casting Part Development
This development project successfully demonstrated the effective application of sand core 3D printing for the rapid prototyping of a highly complex cylinder head casting part. The key learnings and conclusions are as follows:
- Accelerated Development and Cost Reduction: Utilizing 3D printed sand cores for prototype casting parts eliminates the lead time and cost associated with designing and manufacturing hard tooling for core boxes. It enables immediate translation of digital design modifications into physical prototypes, providing invaluable data for validating the casting part design and for optimizing the eventual production tooling.
- Surface Finish and Process Simplification Potential: The fine sand grains used in 3D printing produce cores with inherently dense and smooth surfaces. For this specific casting part application, the internal cavity quality achieved without any refractory coating was equivalent to that of coated conventional cores. This suggests that for certain alloys and casting part geometries, the coating step—a significant process cost and energy consumer—might be optional when using 3D printed cores, further streamlining rapid prototyping.
- Criticality of Core Material Properties: The trial results underscore that not all 3D printing sand materials are equal. Key properties such as tensile strength and high-temperature performance are critical discriminators. Materials with higher green and baked strength, lower thermal expansion, and longer high-temperature durability (like Manufacturer B’s) significantly reduce the risk of core fracture and deformation during pouring, directly contributing to the yield of sound casting parts.
- Demand on Mold System Design: The typically higher and faster gas evolution from 3D printed binder systems necessitates careful attention to mold and core venting design when developing the process for new casting parts. Inadequate venting will lead to gas-related defects, compromising the integrity of the prototype casting parts.
- Integrated Process-Design Optimization: The successful outcome was not solely due to material selection but also involved strategic design-for-manufacturing changes (integrated core printing, gating relocation, venting enhancement) and process adaptations (omission of coating/drying). This highlights the synergistic approach needed in developing casting parts via additive tooling.
In summary, sand core 3D printing has proven to be a powerful and pragmatic tool for the agile development of complex castings like cylinder heads. It shifts the economic equation for low-volume and prototype work, allowing foundries to respond faster to design changes, test new materials like RuT450 for high-performance casting parts, and de-risk the subsequent investment in high-volume production tooling. The technology’s value extends beyond mere prototyping; it serves as a critical bridge between digital design and physical validation, ensuring that final production tooling is developed for a fully optimized and manufacturable casting part.