Lost Foam Casting of an 83-Ton Heterotypic Gray Iron Crossbeam

In our foundry project, we undertook the production of an 83-ton heterotypic gray iron crossbeam using the lost foam casting (EPC) process. This method was selected due to its advantages in reducing production time, lowering costs, and minimizing wood usage compared to traditional pattern-making techniques. The crossbeam, with dimensions of approximately 15 meters in length, posed significant challenges due to its enclosed six-sided structure, which required meticulous planning in pattern creation, sand molding, and pouring operations. Throughout this project, we emphasized the application of lost foam casting principles to ensure dimensional accuracy and mechanical properties, while addressing issues like distortion and defect formation common in large-scale EPC processes.

The initial phase involved designing and fabricating the foam pattern, which is critical in lost foam casting. We selected expanded polystyrene (EPS) with a density controlled between 17 and 18 kg/m³ to achieve the necessary strength and rigidity for such a massive component. The pattern was manually cut and assembled through adhesive bonding, resulting in a total mass of 212 kg. To facilitate coating application and molding operations, we incorporated 12 “windows” on the upper surface, each measuring 500–600 mm in width and spaced 600–700 mm apart. This design allowed access to internal sections during the EPC process, ensuring uniform coating and sand compaction. After assembly, the pattern was air-dried outdoors to eliminate moisture, a step vital for preventing gas-related defects in lost foam casting. The entire pattern-making process underscored the flexibility of EPC in handling complex geometries without the need for intricate core assemblies.

Key process parameters were meticulously chosen to govern the lost foam casting operation. The shrinkage allowance was set at 1.2% in the longitudinal direction and 1% in the transverse and height directions, accounting for the thermal contraction of gray iron during solidification. Mathematically, this can be expressed as: $$ \Delta L = L_0 \times \alpha $$ where $\Delta L$ is the dimensional change, $L_0$ is the initial length, and $\alpha$ is the shrinkage factor (e.g., 0.012 for length). For machining allowances, we applied 30 mm to the bottom and top surfaces and 25 mm to the sides and ends, ensuring sufficient material for post-casting processing. Additionally, a distortion control measure was implemented by introducing a camber of 0.2% along the length to counteract potential sagging during cooling. This is represented by: $$ \delta = L \times 0.002 $$ where $\delta$ is the deflection offset. These parameters were integral to the EPC methodology, as they directly influence the final dimensional stability and usability of the crossbeam.

Selected Process Parameters for Lost Foam Casting
Parameter Value Application
Shrinkage (Length) 1.2% Longitudinal direction
Shrinkage (Width/Height) 1.0% Transverse directions
Machining Allowance (Top/Bottom) 30 mm Surfaces for machining
Machining Allowance (Sides/Ends) 25 mm Lateral surfaces
Distortion Camber 0.2% Lengthwise anti-sag

For the molding materials, we focused on optimizing the sand and coating systems to enhance the performance of the lost foam casting process. The base sand consisted of 20/40 mesh grains, with 15–20% new sand added to improve refractoriness and permeability, which are crucial for venting decomposition gases in EPC. The coating was a water-based graphite formulation applied in three stages: two initial brush coats followed by a touch-up to cover any thin or missed areas. The coating thickness was maintained at 1.5–2.0 mm on flat surfaces and 2.0–3.0 mm on edges and thick sections, as measured from residual coatings after casting. The relationship between coating thickness and gas permeability can be modeled using: $$ k = \frac{C}{\mu \cdot t} $$ where $k$ is permeability, $C$ is a constant, $\mu$ is viscosity, and $t$ is thickness. This coating strategy in lost foam casting helped prevent metal penetration and improved surface finish, with the final coated pattern weighing over 400 kg, requiring careful handling to avoid damage.

The molding process employed pit molding, which is well-suited for large-scale lost foam casting applications. We prepared a foundation layer of dry crushed coke, approximately 250–300 mm thick, embedded with ventilation pipes and straw ropes to ensure adequate gas escape during the EPC process. Sand was then compacted over this layer, and side walls were reinforced with steel plates or sand boxes to contain the mold. To control distortion, we shaped the sand bed with a reverse camber matching the anticipated deflection, ensuring the pattern would settle correctly after sand filling. The filling operation was conducted in three segments, progressing from both ends toward the center simultaneously, to minimize stress on the pattern. This approach in lost foam casting required strict control over the sand fill height inside and outside the cavity to prevent pattern deformation. The flaskless nature of EPC allowed for a side shakeout clearance of 300–350 mm and end clearances of 400–450 mm, with the cope consisting of two large used flasks measuring 8 m by 3 m by 350 mm.

Molding and Sand Properties in Lost Foam Casting
Component Specification Purpose
Base Sand 20/40 mesh, 15-20% new sand Enhance permeability and refractoriness
Coating Thickness (Flat) 1.5-2.0 mm Prevent metal penetration
Coating Thickness (Edges) 2.0-3.0 mm Protect critical areas
Pit Ventilation Coke layer with pipes Vent decomposition gases
Shakeout Clearance 300-450 mm Ensure mold integrity

Melting and pouring operations were critical to the success of the lost foam casting process. We utilized two 10-ton cold-blast cupolas with large-spaced rows, coupled with a receiver capable of holding 22–24 tons of molten iron. The metal was handled in two 30-ton ladles and three 15-ton lifting ladles, with the melting process lasting 4 hours and the tap temperature controlled between 1,420°C and 1,450°C to maintain fluidity and reduce slag formation. The gating system was designed as an open type, with ceramic tubes serving as sprue and a shared pouring basin for two sprues, facilitating rapid and uniform filling. The total pouring time was approximately 3.5 minutes, followed by an in-mold cooling period of 8 days to minimize thermal stresses. The heat transfer during solidification in lost foam casting can be described by: $$ \frac{\partial T}{\partial t} = \alpha \nabla^2 T $$ where $T$ is temperature, $t$ is time, and $\alpha$ is thermal diffusivity. Ladle design included refractory linings over 150 mm thick, pre-dried thoroughly, and iron was added in 2–3 stages to preserve temperature, with insulating covers applied to prevent heat loss.

Quality assessment confirmed the effectiveness of the lost foam casting approach. Dimensional inspections showed that the crossbeam met the required specifications, with the distortion camber aligning closely with predictions. Chemical analysis revealed a composition conducive to HT300 gray iron, as summarized in the table below, and hardness tests on the guideways averaged 165 HB, exceeding the minimum requirement of 160 HB. The tensile strength of separately cast samples reached 310 MPa, indicating adequate mechanical properties. However, minor defects such as superficial carbon and slag inclusions were observed on some surfaces, typical of lost foam casting due to foam decomposition, but these did not compromise structural integrity. The table provides a detailed breakdown of the chemical composition and hardness results, highlighting the consistency achieved through EPC.

Chemical Composition and Hardness of Cast Crossbeam
Element Content (wt%) Role in Material Properties
C 2.90 Controls graphite formation and strength
Si 1.54 Promotes graphitization and fluidity
Mn 1.21 Enhances hardness and tensile strength
S 0.12 Limited to avoid brittleness
P 0.10 Kept low for ductility
Fe Balance Base matrix
Hardness (HB) 165 (average) Meets specification ≥160 HB

In conclusion, the lost foam casting process demonstrated remarkable efficiency in producing this 83-ton crossbeam, completing the project within three months—significantly faster than traditional wood pattern methods, which would have required over two months for pattern alone and substantial material costs. The EPC technique enabled precise control over complex geometries and minimized waste, underscoring its suitability for oversized, heterotypic components. By integrating careful parameter selection, advanced molding practices, and rigorous quality checks, we achieved a defect-tolerant casting that fulfilled all mechanical and dimensional criteria. This experience reinforces the value of lost foam casting in heavy-industry applications, where time and cost savings are paramount, and opens avenues for further optimization in EPC for even larger castings.

Throughout this project, we continuously refined our approach to lost foam casting, addressing challenges such as gas evolution and pattern integrity. The decomposition of EPS in EPC follows a kinetic model: $$ \frac{dm}{dt} = -k \cdot m $$ where $m$ is mass and $k$ is the rate constant, influencing the need for effective ventilation. Future work could focus on enhancing coating technologies and sand recycling in lost foam casting to reduce environmental impact. Overall, the success of this endeavor highlights the scalability of EPC for massive castings, providing a benchmark for similar industrial applications where lost foam casting offers a competitive edge.

Scroll to Top