Throughout my postgraduate research, I have been deeply engaged in the development of a sustainable and cost-effective foundry technology that combines the use of industrial by-products with advanced numerical simulation. The core of my work revolves around the lost foam casting process, a near-net-shape and environmentally friendly casting method, applied to the production of grey iron components using water granulated slag iron as the primary metallic charge. This article presents a comprehensive account of my research, including the novel process methodology, experimental validation of mechanical and metallurgical properties, and the numerical simulation of mould filling and solidification using ProCAST and Visual Environment software. The study systematically investigates different gating system designs, identifies typical casting defects such as shrinkage porosity and slag inclusions, and proposes an improved stepped gating system that effectively mitigates these problems. The findings demonstrate that lost foam casting, when combined with the innovative use of slag iron, can achieve both economic and technical benefits while maintaining product quality.
Introduction to Lost Foam Casting
Lost foam casting (LFC), also known as evaporative pattern casting (EPC), is a casting process that uses a foam pattern made of expanded polystyrene (EPS), STMMA, or EPMMA. The foam pattern, which is an exact replica of the desired casting, is coated with a refractory coating, dried, and then embedded in dry sand within a flask. The sand is compacted by vibration, and a vacuum is applied to stabilize the mould. During pouring, the molten metal vaporizes the foam pattern, and the liquid metal takes its place, forming the casting after solidification and cooling. The process offers numerous advantages over conventional sand casting, including excellent dimensional accuracy, minimal machining allowances, simplified production steps, and the ability to produce complex geometries without cores or parting lines. Because the foam pattern remains in the mould, the process eliminates the need for mould assembly and reduces the risk of misalignment. Furthermore, the use of dry sand without binders makes sand reclamation easier and reduces environmental pollution. These attributes have earned lost foam casting the reputation of being a “21st-century casting technology”.
The history of lost foam casting dates back to 1956 when H.F. Shroyer patented the full mould process. In the 1960s, German researchers introduced the dry sand lost foam method, and the combination with vacuum sealing led to the modern lost foam casting technique. Since the 1980s, the process has been industrialized, particularly in aluminium alloy castings for the automotive industry. However, its application to ferrous castings, especially grey iron, has been more limited due to the higher pouring temperatures and the greater complexity of foam degradation products. Nevertheless, the potential benefits for iron castings are substantial, and my research aims to contribute to this field by exploring the use of a low-cost raw material—water granulated slag iron—combined with process simulation to optimize the lost foam casting of grey iron components.
Water Granulated Slag Iron as a Foundry Raw Material
Water granulated slag iron is a by-product obtained from the blast furnace or electric arc furnace process. When molten slag is quenched in water, it forms a glassy granular material. This slag contains a certain amount of iron oxide and metallic iron particles, which can be recovered through crushing and magnetic separation. The resulting product, known as water granulated slag iron, typically contains a significant proportion of metallic iron and can serve as a substitute for traditional steel scrap and foundry returns in the production of grey iron. The market price of water granulated slag iron is considerably lower than that of conventional charge materials, offering an attractive economic advantage. However, direct melting of slag iron in an induction furnace poses a technical challenge: when heated alone, slag iron particles tend to agglomerate into a lumpy mass that does not melt even at temperatures well above the melting point of iron. This phenomenon has hindered the direct use of slag iron in foundry practice.
To overcome this difficulty, I developed and patented a method for directly producing grey iron castings from water granulated slag iron. The key innovation is to first melt a portion of conventional grey iron charge materials in the furnace. Once a molten pool of iron is established, the slag iron particles are added gradually. The molten iron provides a favorable thermodynamic environment that facilitates the complete dissolution and melting of the slag iron particles. The resulting melt is then treated with refining agents, inoculants, and nodulizers (if required) to achieve the desired chemical composition and microstructure. This method effectively solves the agglomeration problem and enables the utilization of slag iron without any special pre-treatment. The process not only reduces raw material costs by 1000–1500 RMB per ton but also shortens melting time and energy consumption. For a typical 750 kg induction furnace, replacing 400 kg of conventional charge with slag iron saves approximately 100–150 kWh per heat, leading to a daily cost reduction of 4000–4500 RMB when running ten heats per day. This substantial economic benefit makes the process highly attractive for commercial foundries.
Process Description for Lost Foam Casting of Grey Iron from Slag Iron
The complete production route for manufacturing grey iron castings using water granulated slag iron through the lost foam process is illustrated in the following steps. First, the slag iron is obtained from the slag quenching process and crushed to particles of 1–3 mm using a vertical mill and then magnetically separated to achieve a high iron content. Second, a portion of conventional grey iron charge is placed at the bottom of an induction furnace. The furnace is preheated for 10–15 minutes at low power, then the power is increased to full capacity to melt the charge. Once a molten pool is formed at a temperature above the liquidus line, the prepared slag iron particles are added into the melt. The furnace power is maintained at full capacity, and the temperature is raised to 1450–1500°C. The slag iron particles gradually dissolve and melt into the iron melt over a period of about two hours. Third, a sample is taken for rapid chemical analysis using a carbon-silicon analyzer. Adjustments are made to the carbon and silicon contents by adding steel scrap, carburizer, and ferrosilicon as needed to meet the target composition for the desired grey iron grade, such as HT150. The target composition ranges for HT150 are presented in the table below.
| Element | HT150 Target Composition (wt%) |
|---|---|
| C | 3.0 – 3.4 |
| Si | 1.8 – 2.1 |
| Mn | 0.5 – 0.8 |
| P | < 0.2 |
| S | ≤ 0.12 |
Once the composition is adjusted, the melt is heated to a tapping temperature of 1390–1450°C. Before tapping, a mixture of refining agents (e.g., cryolite), modifying agents (e.g., rare earth ferrosilicon), and inoculants (e.g., ferrosilicon FeSi75) is placed in a preheated ladle. The amounts typically range from 0.1% to 0.2% of the melt weight for each component. After tapping and thorough stirring for 1–2 minutes, the melt is ready for pouring. For the lost foam process, the melt temperature at pouring is set to 1450°C to compensate for the heat absorbed by foam degradation and to ensure complete filling of the intricate pattern.
The lost foam patterns for grey iron castings are produced by expanding foam beads in a metal mould. For the specific component studied in this research—a lift base frame with dimensions 770 mm × 640 mm × 60 mm and a mass of about 114 kg—the pattern was manufactured as a single piece. The gating system and risers were made manually from foam board and attached using hot-melt adhesive. A refractory coating with good permeability, such as the Guilin No.5 coating, was applied by dipping. The coating thickness was 1.5–2 mm, applied in two layers with intermediate drying. The coated pattern cluster was then placed in a vented sand box, filled with dry silica sand (or ceramic sand), and compacted by three-dimensional vibration. A vacuum of 0.05 MPa was applied during pouring, and the vacuum was maintained for 10–15 minutes after filling before reducing to 0.025–0.04 MPa during solidification to provide mould rigidity and assist in the removal of degradation gases.
Experimental Validation: Metallography and Mechanical Properties
To verify the quality of grey iron castings produced by this method, I conducted extensive experiments on samples taken from a small bearing housing cast from slag iron via lost foam casting. Metallographic specimens were prepared by standard grinding and polishing techniques, followed by etching with 4% nital solution. The microstructures were observed under an optical microscope and a scanning electron microscope. The results revealed a typical grey iron microstructure consisting of pearlite, ferrite, and graphite flakes. The graphite was predominantly Type A flake graphite, uniformly distributed, with a length of approximately 120–150 μm, which is characteristic of HT150 grade iron. The pearlite content was in the range of 40–90%, with fine lamellar spacing. No abnormal phases or excessive carbides were observed, confirming that the slag iron melting process did not introduce any deleterious microstructural features. The microstructure was essentially identical to that of conventionally melted HT150, as shown in the comparative study.
Mechanical property tests were carried out following the Chinese national standard GB/T 9439-1988. Tensile tests were performed on flat specimens machined from the casting wall with a thickness of 45 mm. The specimen geometry had a gauge length of 28.25 mm, width of 12.5 mm, and thickness of 2 mm. The test was conducted using a universal testing machine at a constant crosshead speed. The tensile stress–strain curve exhibited linear elastic behavior up to fracture, with no significant plastic deformation, which is typical for grey cast iron. The maximum tensile force recorded was 2.888 × 10³ N, resulting in a tensile strength of:
$$ \sigma_b = \frac{F_{\text{max}}}{S_0} = \frac{2.888 \times 10^3 \ \text{N}}{25 \ \text{mm}^2} = 115 \ \text{MPa} $$
According to the standard for grey iron castings, HT150 with a wall thickness between 40 mm and 80 mm requires a minimum tensile strength of 105 MPa. The obtained value of 115 MPa exceeds this requirement, confirming that the slag iron based castings meet the mechanical performance criteria for HT150 grade.
In addition to tensile strength, I evaluated the transverse bending strength using a three-point bending fixture. Rectangular specimens with dimensions of 6.5 mm × 6.5 mm × 30 mm were machined from the same casting, and the support span was set to 28 mm. The bending test was performed on the same universal testing machine. The specimen fractured at a maximum load of 2.13 × 10³ N. The bending strength was calculated using the formula for three-point bending:
$$ \sigma_{bb} = \frac{3 F L}{2 b h^2} = \frac{3 \times 2.13 \times 10^3 \times 28}{2 \times 6.5 \times 6.5^2} = 325.75 \ \text{MPa} $$
The bending strength of 325.75 MPa is substantially higher than the typical range of 150–250 MPa required for HT150, demonstrating excellent load-bearing capacity. These mechanical tests confirmed that the lost foam casting process using water granulated slag iron can produce grey iron components with mechanical properties that fully comply with the HT150 specification. The combination of economic and technical feasibility makes this method a viable alternative for foundries seeking to reduce costs without compromising product quality.
Numerical Simulation of the Lost Foam Casting Process
Numerical simulation plays a crucial role in optimizing the lost foam casting process, particularly for complex geometries and when developing new process routes. In my research, I used ProCAST, a finite element based casting simulation software, together with Visual Environment for pre-processing and mesh generation. The simulation workflow consists of several steps: geometry creation, mesh generation, material property assignment, boundary condition setup, and solver execution.
The geometric model of the casting, gating system, and risers was created in Pro/E and exported in IGES format. The model was then imported into Visual-Mesh, where it was cleaned and repaired. Tetrahedral finite element meshes were generated with different element sizes: a fine mesh of 5 mm for the casting body to capture temperature gradients accurately, a coarser mesh of 20 mm for the gating system and risers, and an even coarser mesh of 50 mm for the sand mould. The mesh model is shown in the figure for both the bottom gating and top gating configurations.
For the lost foam simulation, it is essential to define the material properties for the molten iron, the foam pattern, and the dry sand mould. The grey iron HT150 material data were taken from the ProCAST database and customised with the appropriate liquidus and solidus temperatures, latent heat, and thermal conductivity. The key thermophysical properties used in the simulation are listed below.
| Property | Value |
|---|---|
| Liquidus temperature (°C) | 1130 |
| Solidus temperature (°C) | 1106 |
| Latent heat (kJ/kg) | 241 |
| Pouring temperature (°C) | 1450 |
| Initial mould temperature (°C) | 25 |
| Vacuum pressure (MPa) | 0.05 |
The interface between the molten metal and the foam pattern was set as “EQUIV” to allow continuous heat and mass transfer, while the interfaces between the metal and the sand mould and between the foam and the mould were set as “COINC”. Boundary conditions included a pressure boundary at the top of the pouring cup (representing the atmospheric pressure plus the metallostatic head) and a pressure boundary on the external surfaces of the sand mould (representing the vacuum level). The heat loss from the mould surfaces was modelled using a convection coefficient to the ambient air. The simulation was run in two steps: first, a coupled mould filling and heat transfer analysis using the VOF method to track the free surface of the molten metal; second, a solidification analysis including the effect of graphite expansion on porosity formation.
Governing Equations for Mould Filling
The fluid flow of molten metal during lost foam casting is governed by the Navier-Stokes equations for incompressible Newtonian fluids, together with the continuity equation and the energy equation. The continuity equation is expressed as:
$$ \frac{\partial \rho}{\partial t} + \nabla \cdot (\rho \vec{V}) = 0 $$
For incompressible flow, this simplifies to:
$$ \frac{\partial u}{\partial x} + \frac{\partial v}{\partial y} + \frac{\partial w}{\partial z} = 0 $$
The momentum equations in the three spatial directions are:
$$ \rho \left( \frac{\partial u}{\partial t} + u \frac{\partial u}{\partial x} + v \frac{\partial u}{\partial y} + w \frac{\partial u}{\partial z} \right) = -\frac{\partial P}{\partial x} + \rho g_x + \mu \nabla^2 u $$
$$ \rho \left( \frac{\partial v}{\partial t} + u \frac{\partial v}{\partial x} + v \frac{\partial v}{\partial y} + w \frac{\partial v}{\partial z} \right) = -\frac{\partial P}{\partial y} + \rho g_y + \mu \nabla^2 v $$
$$ \rho \left( \frac{\partial w}{\partial t} + u \frac{\partial w}{\partial x} + v \frac{\partial w}{\partial y} + w \frac{\partial w}{\partial z} \right) = -\frac{\partial P}{\partial z} + \rho g_z + \mu \nabla^2 w $$
The energy conservation equation during mould filling, accounting for the heat absorbed by foam decomposition, is given by:
$$ \rho C_p \left( \frac{\partial T}{\partial t} + u \frac{\partial T}{\partial x} + v \frac{\partial T}{\partial y} + w \frac{\partial T}{\partial z} \right) = \lambda \nabla^2 T + \rho L \frac{\partial f_s}{\partial t} $$
where \( \rho \) is the density, \( C_p \) is the specific heat, \( \lambda \) is the thermal conductivity, \( L \) is the latent heat, and \( f_s \) is the solid fraction.
To track the free surface of the molten metal, the volume of fluid (VOF) method is used. The volume fraction \( F \) satisfies:
$$ \frac{\partial F}{\partial t} + u \frac{\partial F}{\partial x} + v \frac{\partial F}{\partial y} + w \frac{\partial F}{\partial z} = 0 $$
where \( F \) ranges from 0 (empty cell) to 1 (fully filled cell). The lost foam process introduces an additional gas pressure in the gap between the molten metal front and the foam pattern, which can be modelled using a simplified pressure model:
$$ P_{\text{gap}} = P_0 + \frac{\rho_{\text{foam}} L_{\text{foam}} u_{\text{front}}}{K_{\text{coat}} / B_{\text{coat}}} $$
where \( \rho_{\text{foam}} \) is the foam density, \( L_{\text{foam}} \) is the foam decomposition heat, \( u_{\text{front}} \) is the front velocity, and \( K_{\text{coat}} / B_{\text{coat}} \) represents the coating permeability. In the ProCAST setup, these effects are included through the special lost foam boundary conditions and material parameters.
Solidification and Porosity Prediction
Solidification of grey iron is accompanied by the release of latent heat and the formation of graphite. The temperature field during solidification is obtained by solving the Fourier heat conduction equation with a source term for latent heat:
$$ \rho C_p \frac{\partial T}{\partial t} = \frac{\partial}{\partial x} \left( \lambda \frac{\partial T}{\partial x} \right) + \frac{\partial}{\partial y} \left( \lambda \frac{\partial T}{\partial y} \right) + \frac{\partial}{\partial z} \left( \lambda \frac{\partial T}{\partial z} \right) + \rho L \frac{\partial f_s}{\partial t} $$
The latent heat is handled using the enthalpy method, where the total enthalpy \( H \) is defined as:
$$ H = H_0 + \int_{T_0}^{T} C_p \, dT + L (1 – f_s) $$
Porosity formation is predicted using the Niyama criterion, which correlates the thermal gradient \( G \) and cooling rate \( R \) with micro-porosity. The Niyama parameter is:
$$ N_y = \frac{G}{\sqrt{R}} $$
A low Niyama value indicates a high risk of shrinkage porosity. In addition, the graphite expansion during eutectic solidification of grey iron provides internal self-feeding, which can reduce porosity. In the simulation, a graphite expansion factor of 0.5 was initially used, and later adjusted to 0.6 to match experimental results.
Simulation Results for Different Gating Systems
Bottom Gating System (Trial 1)
The bottom gating system was designed with a single ingate at the bottom of the casting. The simulation results showed that the mould filling took 56.14 seconds, which is considerably longer than the filling of an equivalent conventional sand mould due to the back pressure from foam degradation. The filling pattern was smooth and progressive, as the molten metal rose from the bottom to the top of the casting. However, the temperature loss during the long filling time was significant, especially at the farthest regions from the ingate. The final filling regions were located near the risers at the top, where the metal temperature had dropped considerably. This could lead to incomplete decomposition of the foam and increased risk of slag entrapment. The solidification analysis revealed that the risers solidified before the thick sections of the casting, creating isolated liquid pools that were not adequately fed. The resulting porosity distribution, shown in the simulation, displayed prominent shrinkage cavities in the central thick sections. The actual trial casting exhibited similar defects, confirming the accuracy of the simulation.
Top Gating System (Trial 2)
The top gating system allowed faster mould filling, with a total time of 52.14 seconds, because the shorter runner and direct entry of molten metal into the cavity reduced heat loss. However, the filling front progressed from the top downwards, which meant that the slag and foam degradation residues were pushed towards the bottom of the casting. As the flow direction was opposite to the direction of flotation, slag inclusions tended to become trapped in the lower part of the casting. The solidification simulation showed that although the risers at the top were effective in feeding the upper sections, the lower thick sections still suffered from shrinkage porosity due to inadequate feeding paths. This was consistent with the trial casting, which showed extensive slag inclusions and some shrinkage cavities.
Improved Stepped Gating System
Based on the shortcomings of both bottom and top gating, I designed an improved stepped gating system with two layers of ingates. The lower layer filled the bottom part of the casting, while the upper layer filled the upper part, allowing a more uniform temperature distribution and better feeding. The simulation results for the stepped system showed a reduction in total filling time to 41.13 seconds, a 20% improvement over the bottom gating system. The filling was smooth, with the final filling zone located at the risers, which facilitated the entrapment of slag and oxides. Solidification analysis revealed that the upper ingates remained liquid longer, providing effective feeding to the upper hot spots. The porosity distribution, as shown in the simulation, displayed only minor shrinkage in a limited area of the lower thick section. The actual production with the stepped gating system confirmed these predictions: the castings were free of slag inclusions, and only slight dispersed shrinkage remained, well within acceptable limits for HT150 industrial castings. The improved process resulted in castings with excellent surface quality, no cold shuts, and no cracks, meeting all the customer requirements.
Comparison of Simulation and Experimental Results
The following table summarizes the key characteristics of the three gating systems and their corresponding casting quality.
| Gating System | Filling Time (s) | Flling Pattern | Slag Inclusions | Shrinkage Porosity | Overall Quality |
|---|---|---|---|---|---|
| Bottom gating | 56.14 | Smooth, upward | Minimal | Significant | Poor |
| Top gating | 52.14 | Downward | Severe | Moderate | Poor |
| Stepped gating | 41.13 | Progressive, uniform | None | Minor | Good |
The excellent correlation between the simulation predictions and the trial castings validated the numerical model and the selected process parameters. The graphite expansion factor, interface heat transfer coefficients, and vacuum conditions were crucial in matching the solidification behavior and porosity formation. After calibration, the simulation could be reliably used to optimize other castings produced by the slag iron lost foam process.
Conclusion
In this research, I have successfully developed and demonstrated a novel foundry process that directly utilizes water granulated slag iron to produce grey iron castings through the lost foam casting route. The key technical challenge of melting slag iron was overcome by first establishing a molten pool of conventional grey iron, into which the slag iron particles were subsequently added and completely dissolved. This method not only lowers material costs but also reduces melting time and energy consumption, providing a significant economic benefit for foundries. Experimental verification showed that the resulting HT150 castings possess microstructures and mechanical properties equivalent to those produced from conventional charges. The tensile strength of 115 MPa and bending strength of 325.75 MPa exceeded the HT150 standard requirements, confirming the technical viability of the process.
Numerical simulation using ProCAST proved to be an indispensable tool for process development. By simulating the mould filling and solidification of different gating systems, I was able to identify potential defects—such as shrinkage porosity and slag inclusions—and to design an improved stepped gating system that minimized these defects. The simulation results were in good agreement with actual trial castings, thereby establishing a reliable simulation methodology for lost foam casting of grey iron. The optimized process produced castings with sound internal structure and clean surfaces, fully satisfying the industrial application requirements.
The combination of lost foam casting with the utilization of industrial by-products represents a sustainable approach for the foundry industry. The methodology not only conserves resources and reduces environmental impact but also enhances economic efficiency. Future work may extend this approach to ductile iron castings and more complex geometries, as well as to the production of aluminum and magnesium alloy components. The comprehensive understanding of the process parameters, material properties, and simulation techniques gained from this study will serve as a solid foundation for further innovations in lost foam casting technology.