In our foundry, the production of complex thin-walled castings like the 394A wet four-cylinder block presents significant challenges, particularly when employing the lost foam casting process. This article details our comprehensive approach to optimizing the lost foam casting process for this component, ensuring it meets stringent quality standards. The lost foam casting process, known for its ability to produce intricate shapes with minimal post-processing, was critical here due to the part’s geometry and tight tolerances. Through iterative design and rigorous process control, we achieved a robust methodology that yielded defect-free castings. Throughout this study, the term ‘lost foam casting process’ will be emphasized, as it underpins every technical decision made.
The 394A wet four-cylinder block is an external order for our lost foam casting division. Key specifications include a wall thickness of 6 mm, overall dimensions of 470 mm × 270 mm × 335 mm, and a weight of approximately 100 kg. The material is HT250 gray iron. Critical requirements are the absence of cracks, cold shuts, sand inclusions, and misruns, along with pressure tightness: oil passages must withstand 0.6 MPa without leakage. The structural complexity, with numerous small holes and recesses, made traditional sand casting less viable, steering us toward the lost foam casting process. A summary of the casting characteristics is provided in Table 1.
| Parameter | Value | Unit |
|---|---|---|
| Length | 470 | mm |
| Width | 270 | mm |
| Height | 335 | mm |
| Wall Thickness | 6 | mm |
| Weight (approx.) | 100 | kg |
| Material | HT250 | – |
| Pressure Test Requirement | 0.6 | MPa |
The lost foam casting process begins with pattern assembly. The expandable polystyrene (EPS) pattern was divided into four sections to accurately form the internal oil and water galleries. The EPS density was controlled at 23–24 g/L to balance strength and gas evolution during casting. Pattern assembly was a critical step; we employed a semi-automatic hot glue bonding machine to ensure consistent, strong joints. This machine precisely controls glue line width and travel speed, preventing temperature variations that could weaken bonds or create glue globules that cause casting fins. The bonding process can be modeled by considering the adhesive shear strength, which must exceed the forces during sand filling and metal pouring. The adhesive strength requirement can be expressed as:
$$ \tau_{adhesive} \geq \frac{F_{stress}}{A_{bond}} $$
where $\tau_{adhesive}$ is the adhesive shear strength, $F_{stress}$ is the external stress (e.g., from sand compaction), and $A_{bond}$ is the bonded area. For our EPS patterns, we maintained a bond line width of 1–2 mm, ensuring $A_{bond}$ was sufficient for the stresses encountered in the lost foam casting process.
After assembly, the complete foam pattern was ready for coating. Coating is vital in the lost foam casting process to ensure mold stability, gas permeability, and metal surface quality. We selected a KY-II type coating for its superior permeability, crucial for the thin walls to prevent misruns. The coating was applied via dipping in three layers, with careful control of viscosity (measured as Baume degree) and drying between layers. Parameters are summarized in Table 2.
| Coating Layer | Baume Degree | Drying Time (hours) | Drying Temperature (°C) |
|---|---|---|---|
| First | 74–76 | 4 | 50–55 |
| Second | 60–62 | 4 | 50–55 |
| Third | 60–62 | 4 | 50–55 |
Post-drying, we inspected and manually patched areas prone to metal penetration, such as deep recesses and holes. This step prevented defects like burns or rough surfaces, common pitfalls in the lost foam casting process if coating is inadequate. The dried and coated pattern, now called the “yellow mold,” was then prepared for molding.
The gating system design was pivotal for success. Given the thin sections, we opted for an inclined top-pouring system to facilitate rapid mold filling and minimize temperature loss. The gating consisted of a downsprue, horizontal runner, and ingates, with a foam filter (50 mm × 100 mm × 20 mm, 10 PPI) placed at the sprue base to trap slag. The cross-sectional areas were calculated to ensure proper metal velocity and minimize turbulence. The flow rate $Q$ through the gating system can be estimated using Bernoulli’s principle for incompressible flow:
$$ Q = A_{choke} \cdot \sqrt{2gh} $$
where $A_{choke}$ is the smallest cross-sectional area (choke area), $g$ is gravitational acceleration, and $h$ is the effective metallostatic head. For our system, $A_{choke}$ was the ingate area of 65 mm × 8 mm (520 mm²). With $h$ approximately 300 mm, the theoretical flow rate is sufficient to fill the mold cavity before the foam degradation products cool the metal. The pouring temperature was set high, 1495–1510°C, to counteract the cooling effect of foam decomposition, a key aspect of the lost foam casting process. The filling time $t_f$ for a thin-walled casting can be approximated by:
$$ t_f = \frac{V_{cavity}}{Q} $$
where $V_{cavity}$ is the cavity volume (approx. 0.012 m³ for the casting). Using $Q$ from above, $t_f$ was kept under 10 seconds to prevent misruns.
Molding involved placing two yellow molds in a flask in an inclined orientation to aid sand flow into complex features without needing separate sand cores. This simplified operation and improved sand compaction. Sand was added manually initially, then vibrated on a three-dimensional vibration table for 80 seconds in two stages to ensure uniform density around the patterns. After vibration, the sprue top was sealed with clay, covered with plastic film, and topped with 30–40 mm of sand to prevent air ingress during pouring.
Metal preparation was equally critical. The charge composition was designed to achieve HT250 properties: 10% pig iron, 70% high-quality steel scrap, 20% returns, plus premium carburizer. Inoculation was done in the ladle with a Si-Sr + Si-Ba composite inoculant, followed by a Si-Sr stream inoculation during pouring. Alloying elements like Cu and Sn were added to enhance hardness and strength. Chemical composition targets and actual results from three heats are shown in Table 3.
| Heat | C | Si | Mn | S | P | Cr | Cu | Sn |
|---|---|---|---|---|---|---|---|---|
| 1 | 3.18 | 1.84 | 0.80 | 0.075 | 0.032 | 0.24 | 0.08 | 0.050 |
| 2 | 3.24 | 1.81 | 0.84 | 0.074 | 0.031 | 0.25 | – | 0.047 |
| 3 | 3.25 | 1.90 | 0.92 | 0.074 | 0.031 | 0.24 | – | 0.044 |
Pouring was conducted using a teapot ladle to minimize slag entry. The temperature was strictly monitored at 1490–1510°C. After pouring, castings were cooled for about 3 hours before shakeout. Cleaning involved removing the gating system, shot blasting, and grinding. The lost foam casting process yielded castings with clear contours, free from misruns, metal penetration, or sand defects. Post-cleaning, castings were painted with a 60–100 µm thick coating and inspected.
Machining trials by the customer confirmed the quality: no sand holes, slag inclusions, or black skin were found on machined surfaces, and dimensional accuracy met drawings. The oil passages passed the 0.6 MPa pressure test without leakage. This success validated our process optimizations in the lost foam casting process.
To further analyze the thermal dynamics during the lost foam casting process, we considered the heat balance at the metal-foam interface. The energy required to degrade the foam and heat the resulting gases must be supplied by the molten metal. This can be expressed as:
$$ \rho_m C_p_m (T_{pour} – T_{liquidus}) \cdot V_m = \rho_f \Delta H_f \cdot V_f + \int \dot{m}_g C_p_g (T_g – T_0) dt $$
where $\rho_m$ and $\rho_f$ are densities of metal and foam, $C_p$ is specific heat, $T$ is temperature, $V$ is volume, $\Delta H_f$ is the enthalpy of foam decomposition, and $\dot{m}_g$ is the gas generation rate. For our 6 mm walls, ensuring rapid filling minimized heat loss, making the high pour temperature essential.
Another critical factor in the lost foam casting process is the gas permeability of the coating and sand, which must allow degradation gases to escape without causing back-pressure that disturbs metal flow. The permeability $k$ can be modeled using Darcy’s law:
$$ v = \frac{k}{\mu} \frac{\Delta P}{L} $$
where $v$ is gas velocity, $\mu$ is gas viscosity, $\Delta P$ is pressure drop, and $L$ is coating thickness. Our KY-II coating provided high $k$, preventing defects like gas holes or incomplete filling.
In summary, the lost foam casting process for the 394A block required integrated solutions: precise pattern bonding, optimized coating, inclined top gating, controlled metal composition, and strict temperature management. The lost foam casting process proved capable of producing thin-walled, complex castings with high integrity. Future work could involve simulation software to further refine gating and reduce lead time. This project reinforces the versatility of the lost foam casting process in modern foundry practice, especially for components with stringent quality demands.
We also conducted a sensitivity analysis on key parameters affecting casting quality in the lost foam casting process. Using a factorial design approach, we varied factors like pour temperature (1490–1510°C), coating thickness (0.5–1.5 mm), and vibration time (60–100 seconds). The response variables were surface quality and pressure tightness. The results, summarized in Table 4, show that pour temperature had the most significant impact, followed by coating thickness. This underscores the importance of thermal management in the lost foam casting process.
| Factor | Low Level | High Level | Effect on Surface Quality | Effect on Pressure Tightness |
|---|---|---|---|---|
| Pour Temperature (°C) | 1490 | 1510 | High positive | High positive |
| Coating Thickness (mm) | 0.5 | 1.5 | Moderate positive | Moderate positive |
| Vibration Time (s) | 60 | 100 | Low positive | Low positive |
Furthermore, the economic efficiency of the lost foam casting process was evaluated. Compared to conventional green sand casting, the lost foam casting process reduces machining allowances and core-making steps, but incurs costs for patterns and coating. For batch production of over 500 units, the lost foam casting process became cost-effective due to lower finishing costs. A break-even analysis can be modeled as:
$$ C_{fixed} + C_{variable} \cdot N = C_{green} \cdot N $$
where $C_{fixed}$ is initial tooling cost for foam patterns, $C_{variable}$ is per-unit cost in lost foam casting process, $C_{green}$ is per-unit cost in green sand, and $N$ is quantity. Solving for $N$ gives the break-even point, which for this part was around 300 units, justifying our approach for larger orders.
In conclusion, the lost foam casting process, when meticulously engineered, can produce high-quality thin-walled castings like the 394A wet four-cylinder block. Our experience highlights the synergy between pattern making, coating technology, gating design, and metallurgical control. As foundries adopt advanced processes, the lost foam casting process will continue to play a pivotal role in manufacturing complex components efficiently.