In my extensive experience within the foundry industry, specializing in the production of heavy-duty, precision components, the casting of large, multi-cavity, windowless box-structure machine tool castings presents a unique set of challenges. These castings, such as columns, beams, and crossrails, are foundational to high-precision machining centers. To achieve the required mass, rigidity, and superior damping characteristics essential for precision stability, a technique known as sand enclosure casting is often specified. This process involves permanently encapsulating sand cores within the internal cavities of the casting, leaving no outlet for sand removal. The primary technical hurdle lies in safely venting the gases generated from these enclosed cores during pouring and preventing core displacement due to metallostatic pressure. Through systematic development and production validation, my team and I have refined a reliable methodology for this critical process. The following discourse details our approach, focusing on the design and application of specialized chills and supports, supplemented with empirical data, formulas, and standardized specifications.
The essence of sand enclosure casting for machine tool castings lies in creating a monolithic structure where the internal sand acts as a permanent filler. This is fundamentally different from conventional casting where cores are removed. The benefits for machine tool castings are significant: increased weight-to-volume ratio, enhanced dynamic stiffness, and improved vibration damping, all contributing to the accuracy retention of the final machine. However, the trapped core gases and the substantial buoyant forces on the cores during metal pouring pose serious risks of blowholes, mistruns, and core shift, which can compromise the casting’s integrity.
Our successful production runs have involved several large machine tool castings, including vertical columns, overhead crossbeams, and top girders, all in grade HT250 grey iron, with poured weights ranging from 4.5 to 12 tonnes. In each case, the internal geometry consisted of multiple sealed chambers. For instance, one column design featured over ten separate enclosed cavities, with a total enclosed sand mass exceeding 5000 kg. The gating system was carefully designed as a stepped or bottom-up rain gate system to ensure smooth, controlled filling and minimize turbulence. However, the cornerstone of the process was the implementation of two custom-designed elements: venting chills and anti-buoyancy machining chills.
The most critical innovation in our sand enclosure process is the venting chill. Since the core has no conventional vent path to the mold exterior, gases must escape through the casting wall itself. The venting chill is a precision-cast, tubular insert made from the same material as the parent casting (e.g., HT250). It is strategically placed at the highest points of the enclosed core, penetrating the casting wall. Its function is dual: it acts as a mechanical support and, more importantly, provides a dedicated conduit for core gases to escape into the mold’s venting system during solidification. The basic design principle is summarized below, and a standard series we developed is tabulated.
The fundamental geometry of a venting chill involves a flange that sits against the mold wall and a hollow stem that penetrates into the core. The length embedded within the core ensures a gas-tight seal with the core sand. The key dimensions are the outer diameter (D), which influences venting capacity, and the internal passage dimensions (d, d0). The venting capacity, or the volumetric gas flow rate \( Q_v \), can be approximated by considering the chill as a short pipe during the gas generation phase:
$$ Q_v \approx \frac{\pi (d_{eff})^4 \Delta P}{128 \mu L} $$
where \( d_{eff} \) is an effective diameter of the gas passage, \( \Delta P \) is the pressure differential between the core and the mold, \( \mu \) is the dynamic viscosity of the hot gas mixture, and \( L \) is the effective length of the chill passage. In practice, we select the chill size based on the volume of the enclosed core. A general rule we established is that the total venting area provided by all chills on a core should be proportional to the core’s volume and the anticipated gas evolution rate of the core sand binder. Our standard series is designed to cover a range of core volumes.
| Chill Designation | Flange Diameter, D (mm) | Major Bore, d (mm) | Minor Bore, d0 (mm) | Flange Thickness, a (mm) | Stem Embedment, b (mm) | Recommended Core Volume Range (m³) |
|---|---|---|---|---|---|---|
| Vent-1 | 60 | 34 | 20 | 4 | 2 | < 0.05 |
| Vent-2 | 80 | 42 | 25 | 6 | 3 | 0.05 – 0.15 |
| Vent-3 | 100 | 50 | 30 | 8 | 4 | 0.15 – 0.30 |
| Vent-4 | 120 | 60 | 35 | 10 | 5 | > 0.30 |
Placement is paramount. We typically use 2 to 4 venting chills per enclosed core, located at the uppermost sections. To ensure accurate positioning, the core box is fitted with location blocks that mirror the chill flange. During coremaking, a dummy rod is inserted into this block to create the precise cavity in the core. After stripping, the venting chill is seated into this cavity, and the surrounding area is carefully packed and coated. Correspondingly, the pattern has this location marked to ensure the mold sand is properly carved out to create a vent channel from the chill’s outer face to the mold’s external venting network. Post-casting, the chill’s bore is sealed according to the surface requirement: if it is on a machined face, it is plugged to relevant standards; if on an as-cast surface, it is welded shut and ground flat.
While venting chills address gas evolution, the substantial buoyant force on the enclosed core must be counteracted to prevent floating and dimensional inaccuracy. The buoyant force \( F_b \) acting on a submerged core is given by Archimedes’ principle:
$$ F_b = \rho_{metal} \cdot g \cdot V_{core} – \rho_{sand} \cdot g \cdot V_{core} \approx \rho_{metal} \cdot g \cdot V_{core} $$
where \( \rho_{metal} \) is the density of molten iron (~7000 kg/m³), \( g \) is gravitational acceleration, \( V_{core} \) is the volume of the enclosed core, and \( \rho_{sand} \) is the core sand density, which is negligible in comparison. For a core of several cubic meters, this force can be on the order of tens of thousands of Newtons. Additionally, there is a dynamic pressure force during mold filling. Ordinary sheet metal chills are entirely inadequate for such loads at molten metal temperatures. Therefore, we employ specially designed cast iron machining chills, which become integral parts of the casting and are later machined away in finished machine tool castings.
These anti-buoyancy chills are designed with a robust cross-sectional area to withstand compressive stress at high temperatures. Their required yield strength at temperature must exceed the stress imposed by the buoyant force distributed over them. The number of chills required, \( n \), is determined by balancing the total buoyant force (with a safety factor) against the cumulative high-temperature compressive strength of the chills and the core sand’s own hot strength. The core sand’s resistance is often the limiting factor. Our derived working formula is:
$$ n \geq \frac{k \cdot \rho_{metal} \cdot g \cdot V_{core}}{A_c \cdot \sigma_c + A_{cs} \cdot \sigma_{sand}} $$
where:
\( k \) = safety factor (1.2 to 1.5)
\( A_c \) = load-bearing cross-sectional area of one chill (m²)
\( \sigma_c \) = high-temperature compressive yield strength of the chill material (Pa)
\( A_{cs} \) = effective contact area between the core and the mold roof per chill (m²)
\( \sigma_{sand} \) = high-temperature compressive strength of the core sand (Pa)
In practice, for the resin sand systems we use, \( \sigma_{sand} \) is relatively low at pouring temperatures, so the chill area \( A_c \) and its high-temperature strength become dominant. We have standardized a series of chills with dimensions optimized for this purpose. The chills are typically “I-beam” or “stud” shaped to provide good anchorage in both the core and the solidified metal.
| Chill Designation | Height Range, H₀ (mm) | Head Diameter, D (mm) | Stem Diameter, d (mm) | Head Thickness, a (mm) | Flange Width, b (mm) | Approx. Load Area, A_c (mm²) |
|---|---|---|---|---|---|---|
| AB-1 | 8 – 20 | 20 | 8 | 1 | 2 | ~50 |
| AB-2 | 15 – 30 | 30 | 10 | 2 | 2 | ~120 |
| AB-3 | 25 – 50 | 40 | 15 | 3 | 3 | ~300 |
| AB-4 | 40 – 80 | 50 | 20 | 5 | 3 | ~500 |
| AB-5 | 60 – 100 | 60 | 25 | 6 | 4 | ~700 |
| AB-6 | 80 – 150 | 70 | 30 | 10 | 5 | ~1100 |
For a large core with a volume \( V_{core} = 2.5 \, \text{m}^3 \), the buoyant force \( F_b \approx 7000 \cdot 9.81 \cdot 2.5 \approx 171,675 \, \text{N} \). Using a safety factor \( k = 1.3 \), and selecting AB-5 chills with \( A_c \approx 7 \times 10^{-4} \, \text{m}^2 \) and estimating \( \sigma_c \) at pouring temperature to be around \( 20 \, \text{MPa} \), and neglecting sand strength for a conservative estimate, the number required would be \( n \geq (1.3 \cdot 171675) / (7e-4 \cdot 20e6) \approx 16 \). We would typically distribute this number evenly across the core’s top surface, often in conjunction with the venting chills.
The overall casting methodology for these large machine tool castings follows sound engineering principles for heavy-section grey iron. Gating systems are designed to minimize erosion and promote directional solidification towards feeders. For sand-enclosed castings, we often prefer bottom-gating or stepped gating to reduce turbulence and initial impact on the cores. Riser design follows the modulus method, compensating for the slower cooling induced by the enclosed sand. However, a unique aspect of sand enclosure casting is the significantly extended cooling time within the mold. The encapsulated sand acts as an excellent insulator, drastically reducing the heat extraction rate from the internal sections of the casting. This must be accounted for in the shakeout schedule to avoid casting damage due to high-temperature handling or the development of excessive residual stresses.
We have correlated the required mold retention time, or保温时间, to the casting’s sectional modulus and enclosed sand mass. A generalized guideline, validated through thermal analysis and practical experience, is presented below. The cooling time \( t \) for a sand-enclosed casting can be modeled by a simplified lumped-capacitance approach modified for a composite wall (metal + sand):
$$ t \propto \frac{(\rho C_p V)_{casting} + (\rho C_p V)_{sand}}{h_{eff} A_{surface}} \ln \left( \frac{T_{pour} – T_{mold}}{T_{shakeout} – T_{mold}} \right) $$
where \( \rho C_p V \) terms represent the thermal capacitance of the casting and the enclosed sand, \( h_{eff} \) is an effective heat transfer coefficient for the mold/sand system, \( A_{surface} \) is the external surface area of the casting, and \( T \) values are temperatures. Since the sand volume \( V_{sand} \) is substantial, the total thermal mass is much larger than a hollow casting. Our practical table for minimum mold retention time before shakeout is based on the total poured weight, which serves as a proxy for thermal mass.
| Total Casting Poured Weight (tonnes) | Minimum Mold Retention Time (hours) | Key Rationale |
|---|---|---|
| 1 – 2 | 24 – 48 | Allows temperature equalization; prevents thermal shock to thick sections. |
| 2 – 5 | 48 – 72 | Ensures pearlitic transformation is largely complete in critical sections. |
| 5 – 10 | 72 – 120 | Manages very high thermal mass; avoids high stress from premature constraint removal. |
| 10 – 20 | 120 – 168 | Essential for ultra-large machine tool castings to reach safe handling temperature throughout. |
Quality control in producing these machine tool castings involves several specific checks. Prior to pouring, each venting chill’s alignment and clear passage to the external vents are verified. The placement and count of anti-buoyancy chills are audited against the calculated requirements for each core. Non-destructive testing, particularly ultrasonic testing, is crucial after cleaning to check for possible shrinkage or gas-related defects near the enclosed areas or chill interfaces. The integrity of the final machine tool castings is paramount, as any internal flaw could lead to vibration or deformation during precision machining operations.
In conclusion, the sand enclosure casting process, when executed with meticulous attention to core venting and fixation, is a highly effective method for producing the massive, rigid, and damped structures required for modern high-performance machine tool castings. The development of standardized venting and anti-buoyancy chill systems, governed by the principles and formulas discussed, has transformed this challenging specification into a reliable and repeatable foundry practice. Our first-hand experience in producing dozens of such castings, with total weights exceeding a hundred tonnes, confirms that the strategic use of these engineered components—coupled with controlled gating, feeding, and extended cooling times—ensures the delivery of sound, high-quality castings. This capability is fundamental to advancing the manufacturing of next-generation, precision machine tools where structural integrity and dynamic performance are non-negotiable. The continuous refinement of these parameters remains a key focus in our pursuit of excellence in casting heavy-section machine tool castings.