The production of large and medium-sized semi-enclosed castings, such as various box-type structures, presents unique challenges within the broader scope of the lost foam casting process. These components, characterized by their thin, uneven walls forming complex internal cavities, are pivotal in machinery for load-bearing and containment functions. While the lost foam casting process offers significant advantages for such geometries—eliminating cores, simplifying production, and enabling near-net-shape forming—it also introduces specific failure modes like mold collapse (sand crushing) and casting expansion (swelling). This article details our first-hand investigation into the root causes of these defects and the practical, effective measures we developed and implemented to significantly improve production yield.
Initial Process Design and Implementation
Our standard lost foam casting process for these components was meticulously designed, covering every stage from pattern making to pouring. The foam patterns were fabricated from expandable polystyrene (EPS) with a density of approximately 10 kg/m³. The gating and risering system was attached manually to ensure integrity. A critical step was the application of a refractory coating. We employed a water-based coating, applying it in multiple layers with specific drying intervals to build adequate thickness and strength without causing pattern distortion. The coating must satisfy multiple requirements: providing a barrier between the metal and sand, allowing gases from the decomposing pattern to escape, and maintaining structural integrity during pouring. The drying parameters are summarized in Table 1.
| Coating Layer | Drying Time (h) | Drying Temperature Range (°C) | Remarks |
|---|---|---|---|
| First Layer | 20 | 35 – 45 | Patterns are repositioned between layers to prevent sagging. Coating must be smooth and uniform, with special attention to corners and recesses. |
| Second Layer | 24 | 40 – 50 | |
| Third Layer | 26 | 45 – 50 |
The molding process involved placing the coated cluster in a flask and surrounding it with unbonded silica sand. The sand was introduced in stages while the flask was vibrated to achieve maximum compaction around the complex geometry. The total vibration time was not less than 600 seconds at a frequency of 40-50 Hz. The final step was to cover the pattern with a plastic film and apply a vacuum to the flask, compacting the sand further and establishing a pressure differential crucial for the process.
For melting, we used selected steel scrap to achieve the target chemistry, as shown in Table 2. The melt was superheated to approximately 1650°C, allowed to settle for slag removal, and then poured at a temperature between 1550°C and 1600°C. The vacuum was maintained during and for at least 10 minutes after the pour.
| Element | C | Si | Mn | P | S | Cr | Ni | Mo | Cu | V |
|---|---|---|---|---|---|---|---|---|---|---|
| Content (max %) | 0.25 | 0.40 | 1.00 | 0.035 | 0.035 | 0.40 | 0.35 | 0.15 | 0.60 | 0.05 |
Analysis of Prevalent Casting Defects
Despite the controlled process, the initial yield for these semi-enclosed castings was unacceptably low, around 40%. The primary failure modes were mold collapse (“塌箱”) and casting expansion/swelling (“涨箱”). Dimensional analysis of defective parts also revealed significant inward distortion of the cavity walls, as evidenced by the measurements in Table 3.
| Feature Measured | Drawing Dimension (mm) | As-Cast Dimension (mm) | Deviation (mm) |
|---|---|---|---|
| Total Casting Length | 1320 | 1308 | -12 |
| Total Casting Width | 980 | 954 | -36 |
| Internal Cavity Length | 1400 | 1386 | -14 |
| Internal Cavity Width | 900 | 870 | -30 |
A systematic root-cause analysis traced these defects to a fundamental imbalance of forces during metal pouring, inherent to the geometry when using the standard lost foam casting process. The core issue was pressure differential. In a semi-enclosed cavity, the sand core is only supported by the vacuum drawn through the surrounding sand mass. If the vacuum level inside the cavity ($P_{vac,int}$) is significantly lower than the pressure outside the cavity ($P_{vac,ext}$), or if the internal sand compaction is poor, a net inward force acts on the walls.
The forces at play can be summarized by considering the pressure balance on the cavity wall. The metal head pressure ($P_{metal}$) and the gas pressure from decomposing foam ($P_{gas}$) act to push the wall outward. Opposing this is the strength of the sand ($\sigma_{sand}$) and the effective external pressure, primarily from the vacuum ($P_{vac,ext}$). Collapse or distortion occurs when:
$$P_{metal} + P_{gas} > \sigma_{sand} + (P_{atm} – P_{vac,ext})$$
Conversely, if the internal cavity lacks sufficient vacuum ($P_{vac,int}$ is too high, closer to atmospheric), the sand within the cavity is poorly compacted and offers little resistance ($\sigma_{sand,int} \approx 0$). The pressure from the metal and foam gases then easily displaces this loose sand, causing the wall to bulge inward—manifesting as the observed swelling and dimensional distortion. The root cause was therefore identified as an inability to establish and maintain an adequate, uniform vacuum within the internal cavity of the casting during the critical pouring and foam decomposition phase.
Developed Solutions and Process Modifications
To rectify this vacuum imbalance, we developed and tested two interconnected modifications to the standard lost foam casting process.
1. Inclined Molding Orientation:
We abandoned the conventional horizontal placement of the casting. Instead, the pattern was positioned in the flask with its open face tilted at a specific angle. This served a crucial purpose: to facilitate the natural flow and effective compaction of unbonded sand into the deep, semi-enclosed cavity. After trials with angles of 21°, 30°, 45°, and 60°, the 45° tilt was found optimal. At 21°, the “top” section had insufficient sand cover. At 30° and 60°, the geometry created pockets where sand flow was hindered, leading to inconsistent compaction and subsequent local defects. The 45° orientation provided a uniform path for sand entry and allowed for effective vibration compaction throughout the cavity volume, significantly improving the intrinsic strength ($\sigma_{sand,int}$) of the internal sand core. The vibration time required for adequate compaction can be considered a function of cavity volume and opening geometry:
$$t_{vibration} = k \cdot \frac{V_{cavity}}{A_{opening}}$$
where $k$ is a constant dependent on sand characteristics and vibration parameters. Inclining the pattern effectively increases the projected opening area $A_{opening}$, reducing the time or energy needed for complete compaction.
2. Implementation of an Internal Vacuum Manifold (External Negative Pressure Tube):
While inclined molding improved sand density, it did not fully solve the vacuum communication problem. To directly address the low vacuum level inside the cavity ($P_{vac,int}$), we designed an internal manifold—a perforated pipe inserted into the cavity during molding. This pipe was connected to the flask’s vacuum system via an external hose. This created a direct path to apply vacuum suction *inside* the cavity, independent of the flow through the sand.
The manifold was constructed from a steel pipe with a 150 mm internal diameter. Holes 2 mm in diameter were drilled in a staggered pattern (15 mm transverse pitch, 10 mm longitudinal pitch) and covered with a wire mesh to prevent sand ingress. This design ensured uniform vacuum draw across the internal volume. During pouring, both the standard flask vacuum (acting externally) and the internal manifold vacuum were activated simultaneously. This dual-system approach was key to balancing the pressure differential across the cavity wall. The goal was to make $P_{vac,int} \approx P_{vac,ext}$, thereby nullifying the net pressure difference that caused distortion. The modified pressure balance equation becomes:
$$P_{metal} + P_{gas} \leq \sigma_{sand} + (P_{atm} – P_{vac,ext}) + (P_{atm} – P_{vac,int})$$
With $P_{vac,int}$ now actively controlled and raised close to $P_{vac,ext}$, the right side of the inequality is substantially increased, stabilizing the cavity walls.
Results, Analysis, and Conclusions
The combined implementation of a 45° inclined molding orientation and the internal vacuum manifold resulted in a dramatic improvement in casting quality. We conducted a production trial of 48 castings (12 batches of 4). The results, compared to the initial process, are summarized in Table 4.
| Process Stage | Defect Type | Initial Process Rate (~40% overall yield) | Modified Process Result | Qualified Rate per Defect |
|---|---|---|---|---|
| Molding & Compaction | Sand Bulge/Inadequate Compaction | High Incidence | 2 defects in 48 castings | ~96% |
| Pouring & Stabilization | Mold Collapse / Casting Swell | Primary failure mode | 3 defects in 48 castings | ~94% |
| Overall Casting Yield | ~90% | |||
The success of the internal manifold validates the pressure differential theory. By providing a dedicated path for vacuum, we ensured the gaseous products from the decomposing foam within the cavity were swiftly evacuated. This prevented a buildup of pressure ($P_{gas}$) and maintained a high $P_{vac,int}$. Consequently, the sand core remained stable against the metallostatic pressure. The manifold is a reusable tool, adding minimal cost or complexity to the lost foam casting process while solving a critical technical hurdle.
In conclusion, the effective production of large, semi-enclosed castings via the lost foam casting process is highly sensitive to pressure dynamics within internal cavities. Standard practice often leads to vacuum deficiency in these zones, causing collapse, swelling, and distortion. Our research and practice demonstrate that:
- Orienting the pattern at a strategic angle (45° for our geometry) is essential for achieving uniform and dense sand compaction in deep cavities.
- The integration of an internal vacuum manifold is a highly effective solution to actively control the pressure environment inside the cavity, balancing forces across the thin walls during metal pouring.
This two-pronged approach directly addresses the core physical challenges, transforming the lost foam casting process into a reliable and high-yield method for producing complex semi-enclosed components, unlocking its full potential for structural castings in industrial applications.