Lost foam casting, often described as the “casting technology of the 21st century,” has been widely recognized for its near-net-shape capability, design flexibility, and environmental cleanliness. However, one of the most stubborn technological bottlenecks that hinders its broader application is the so-called “carbon defect,” which is particularly severe in iron castings. In this paper, I present my systematic investigation of the formation mechanism, influencing factors, and prevention measures for carbon defects in lost foam casting iron castings. Through extensive experiments, advanced analysis, and production practice, I have established a comprehensive understanding that can significantly reduce scrap rates and improve the economic viability of the process.
The work reported here is based on my direct involvement in the lost foam casting workshop of a major foundry, where I have spent years tackling the carbon defect problem. I have used standardized test castings, controlled raw materials, and a wide range of process parameters to isolate the key variables. My goal is to provide a practical and theoretical framework that can help foundry engineers eliminate or minimize carbon defects in their own production.
The fundamental principle of lost foam casting is that a foam pattern is embedded in dry unbonded sand, and molten metal is poured directly onto the pattern. The foam vaporizes as the metal advances, leaving the metal to take the shape of the pattern. The process is shown schematically in the following illustration.
The critical difference from conventional sand casting is that the pattern is not removed before pouring; it is consumed by the heat of the metal. This introduces a complex interaction between the thermal decomposition products of the foam and the advancing metal front. When the decomposition products cannot escape quickly enough through the coated pattern into the surrounding sand, they may become trapped in the metal and form carbonaceous inclusions, commonly known as carbon defects.
Carbon Defects in Lost Foam Casting
Carbon defects appear as black, irregularly shaped inclusions on the surfaces or internal sections of castings. In my experience with lost foam casting iron castings, these defects can be classified into two major types: surface carbon films or wrinkles, and internal carbonaceous inclusions. Both are detrimental to the mechanical properties, pressure tightness, and machinability of the final product.
In a statistical study I conducted over a six-month period in our foundry, I analyzed the scrap reasons for all lost foam casting iron castings. The results are summarized in Table 1.
| Defect Type | Percentage of Total Scrap |
|---|---|
| Carbon defects | 38% |
| Sand inclusion | 22% |
| Shrinkage porosity | 15% |
| Cold shut / misrun | 12% |
| Surface roughness / penetration | 8% |
| Other | 5% |
It is evident that carbon defects constitute the largest single source of scrap in lost foam casting iron production. This motivated me to launch a focused research program aimed at understanding and controlling this defect.
Characterization of Carbon Defects
To characterize the carbon defects, I first cut samples from defective castings. The macro-fracture surfaces showed black, flaky, and sometimes glossy areas, clearly distinct from the gray iron matrix. The boundary between the defect and the normal metal was usually sharply curved, indicating that the carbon-rich material had accumulated at the final solidification front.
Using optical microscopy at 100× magnification, I compared the microstructure of a normal region with that of a defective region. The normal region exhibited uniformly distributed type A graphite flakes in a pearlitic matrix. In contrast, the defective region contained large, massive, and agglomerated graphite clusters that destroyed the continuity of the metallic matrix. These clusters were surrounded by broken and abnormal graphite shapes, which severely degraded the mechanical integrity of the casting.
Scanning electron microscopy (SEM) at 5000× magnification revealed that the normal region had well-defined metal grains and grain boundaries, whereas the defect region showed almost no recognizable metallic grain structure. The defect area appeared as a coherent mass of carbon-rich material.
Energy dispersive X-ray spectroscopy (EDS) line scans across the defect boundary showed a dramatic increase in the carbon peak within the dark regions, while the lighter metallic regions contained typical iron and silicon peaks. Quantitative chemical analysis using an X-ray fluorescence (XRF) analyzer gave the following average compositions for different regions of a defective casting (Table 2).
| Element | Normal Region (wt.%) | Defect Periphery (wt.%) | Defect Core (wt.%) |
|---|---|---|---|
| C | 3.45 | 6.82 | 84.3 |
| Si | 1.85 | 1.62 | 0.51 |
| Mn | 0.62 | 0.58 | 0.12 |
| P | 0.04 | 0.05 | 0.02 |
| S | 0.06 | 0.07 | 0.03 |
| Fe | balance | balance | ~15 |
| O | 0.01 | 0.02 | 0.01 |
The core of the defect is essentially elemental carbon, with only small amounts of iron and impurities. The absence of significant oxygen indicates that the defect is not an oxide inclusion but rather pyrolytic carbon produced by the thermal decomposition of the foam pattern under oxygen-deficient conditions.
Effect of Carbon Defects on Properties
Carbon defects have a pronounced adverse effect on the mechanical properties of lost foam casting iron castings. To quantify this, I prepared tensile test bars using two different pattern materials: an imported copolymer and a domestic polystyrene. The test bars were 30 mm in diameter and were poured together in a single runner system. After machining to standard dimensions, I performed tensile tests. The results are shown in Table 3.
| Bar No. | Pattern Material | Defect Condition | Ultimate Tensile Strength (MPa) |
|---|---|---|---|
| 1 | Copolymer | No visible defect | 245 |
| 2 | Copolymer | No visible defect | 250 |
| 3 | Copolymer | No visible defect | 248 |
| 4 | Polystyrene | Severe carbon defect | 172 |
| 5 | Polystyrene | Light carbon defect | 201 |
| 6 | Polystyrene | Severe carbon defect | 165 |
The presence of carbon defects reduced the tensile strength by up to 32% compared with sound castings. Hardness measurements were also affected. Using a Brinell hardness tester, I measured three points on a defective area and three points on an adjacent normal area. The average hardness in the defective zone was 156 HB, whereas the normal zone measured 187 HB. This indicates a significant softening and degradation of the metallic matrix due to the carbon-rich inclusions.
Machinability tests were performed on a vertical lathe and a machining center. The cutting time for operations crossing a defective area increased by 60%–80% compared with normal regions. Tool wear was accelerated, and the surface finish was poor, often leading to scrapped parts due to unacceptable dimensional accuracy or surface quality. Pressure-tightness tests using kerosene on castings with internal carbon defects showed that after 24 hours, 11 of 13 test castings exhibited leakage through the defect regions. The severity of leakage correlated with the depth and extension of the carbon defect.
Experimental Setup and Raw Materials
To investigate systematically the influence of process variables on carbon defects in lost foam casting, I designed a series of controlled experiments. The main equipment included a pre-expander, a steam autoclave molding machine, a pattern gluing machine, a vibrating table, a self-made coating mixer, and an induction melting furnace. I used a standardized test casting: a front-end housing for a forklift transmission, with a weight of approximately 110 kg, overall dimensions of 460 mm × 400 mm × 380 mm, and a minimum wall thickness of 12 mm. This casting has many machined surfaces and tight dimensional requirements, making it highly sensitive to carbon defects.
Four different pattern materials were selected:
– Domestic expandable polystyrene (EPS) beads from Hangzhou
– Imported EPS beads from Mitsubishi Chemical
– A mixture of 50% imported EPS and 50% domestic EPS
– A copolymer of EPS and polymethyl methacrylate (PMMA), also from Mitsubishi
The basic properties of these materials are listed in Table 4.
| Property | EPS | PMMA | EPS/PMMA Copolymer |
|---|---|---|---|
| Carbon content (wt.%) | 92 | 60 | 77 |
| Specific heat (kJ/kg·K) | 1.4 | 1.2 | 1.3 |
| Decomposition heat (kJ/kg) | −240 | −180 | −205 |
| Glass transition temp. (°C) | 100 | 105 | 103 |
| Initial gasification temp. (°C) | 240 | 260 | 250 |
| Massive gasification temp. (°C) | 420 | 380 | 400 |
Four different coatings were also evaluated:
– A commercial coating from Ashland (USA)
– A commercial coating from Hüttenes-Albertus (Germany)
– A coating developed by Tsinghua University
– A coating jointly developed by Hefei University of Technology and our foundry
Their high-temperature permeability, density, and other key parameters are summarized in Table 5.
| Coating | High-temp. permeability (cm⁴/g·min) at 1000°C | Density (g/cm³) | Remarks |
|---|---|---|---|
| Ashland | 0.35 | 1.45 | High burn-out, good permeability |
| Hüttenes-Albertus | 0.28 | 1.52 | Moderate permeability |
| Tsinghua | 0.33 | 1.48 | Good permeability, poor wettability |
| Self-made (HF) | 0.32 | 1.50 | Improved wettability after modification |
The sand used was quartz sand from Fujian Pingtan, with SiO₂ content of 97.5%. Three different sand size distributions were tested: 20–30 mesh, 30–40 mesh, and 40–50 mesh. Adhesive for pattern assembly included a hot-melt adhesive and a cold-setting adhesive supplied by Hüttenes-Albertus.
Experimental Results and Influencing Factors
Effect of Pattern Material on Carbon Defects
In the first series of tests, I prepared three identical transmission housing patterns from the three different materials: domestic EPS (material A), imported EPS (material B), and a 50/50 mixture (material C). All patterns were coated with the same coating, assembled in the same sand, and poured at 1480°C under the same vacuum level. The results are presented in Table 6.
| Pattern Material | Filling Time (s) | Carbon Defect Severity |
|---|---|---|
| Domestic EPS | 125 | Severe defects on top surface, machining exposed defects 12–18 mm deep |
| Imported EPS | 72 | No visible defects |
| 50/50 Mixture | 95 | Point-like defects only |
The domestic EPS produced long filling times and a large number of liquid decomposition products, which accumulated at the metal front and created a back-pressure that eventually halted the flow. The imported EPS, having a slightly lower carbon content and a more uniform polymeric structure, decomposed more completely into gases, which escaped through the coating easily. The mixture gave intermediate behavior.
The molecular structures of EPS and PMMA are fundamentally different. EPS contains a benzene ring with six carbon atoms, giving a carbon content of about 92%. PMMA contains only five carbon atoms and two oxygen atoms, giving a carbon content of about 60%. In an oxygen-starved casting environment, EPS undergoes random chain scission, yielding large amounts of viscous liquid and solid residues. In contrast, PMMA decomposes via a “unzipping” mechanism, producing mostly gaseous monomers, which are easier to remove. Therefore, using copolymers with a lower carbon content is the most effective way to reduce the source of carbon defects in lost foam casting.
Effect of Pattern Density
I varied the foam pattern density from 0.018 g/cm³ to 0.030 g/cm³ while keeping all other parameters constant. The coating thickness was 2 mm, pouring temperature was 1460°C, and vacuum pressure was 0.05 MPa. The results are shown in Table 7.
| Pattern Density (g/cm³) | Filling Time (s) | Carbon Defect Condition |
|---|---|---|
| 0.018 | 55 | No defects |
| 0.021 | 62 | No defects |
| 0.024 | 70 | Black spots on surface |
| 0.027 | 78 | Clear carbon stains |
| 0.030 | 85 | Severe carbon defects |
Higher pattern density means more mass per unit volume of foam, and consequently more decomposition products must escape through the coating. I therefore recommend using the lowest possible pattern density that still provides adequate stiffness to the pattern assembly. For castings like the transmission housing, an optimum density of 0.020–0.022 g/cm³ was determined.
Effect of Pouring Temperature
Pouring temperature is one of the most sensitive parameters for carbon defect formation. I conducted five groups of experiments using imported EPS patterns and the same coating, sand, and vacuum conditions. The pouring temperature was varied from 1320°C to 1520°C. The results are summarized in Table 8.
| Group | Pouring Temperature (°C) | Results |
|---|---|---|
| 1 | 1320 | Misrun, incomplete filling, severe defects |
| 2 | 1360 | Cold shut, severe carbon defect in the upper region |
| 3 | 1400 | Surface carbon films and internal carbon spots |
| 4 | 1440 | Slight carbon defect at one window after machining |
| 5 | 1480 | No carbon defects |
| 6 | 1520 | Sand fusion on surfaces, but no carbon defect |
Higher temperatures promote the cracking of polystyrene into smaller volatile molecules, which escape easily. I also verified this using a quartz tube sample collector placed in the gap between the metal front and the foam. Gas chromatography results from samples taken at different metal temperatures showed that at 700°C only 27% of the decomposition products were small-molecule gases, whereas at 1450°C the fraction rose to 68%, and at 1600°C to 85%. The remaining products were tar-like liquids and solid residues that cause carbon defects.
In production, for thin-wall castings like the transmission housing, the pouring temperature must be at least 1440°C, and preferably 1480°C. However, excessively high temperatures above 1540°C can cause expansion of the decomposition gas so rapid that it ruptures the coating, leading to sand penetration and metal penetration defects.
Effect of Coating Permeability and Wettability
The coating acts as the pathway through which all decomposition products must pass. Therefore, its high-temperature permeability and its wettability by liquid polystyrene are crucial. In my comparison of the three coatings, I kept the same pattern material (imported EPS), same coating thickness, same sand, and same pouring temperature (1460°C). The results are shown in Table 9.
| Coating | High-temp. Permeability (cm⁴/g·min) | Carbon Defect Rate (%) | Remarks |
|---|---|---|---|
| Ashland | 0.35 | 8 | Good result in machining |
| Hüttenes-Albertus | 0.28 | 18 | Moderate result |
| Self-made (initial) | 0.32 | 27 | Poor wettability caused defects |
| Self-made (modified) | 0.34 | 10 | After adding wetting agents |
Despite the self-made coating having permeability similar to Ashland, its carbon defect rate was much higher. The reason was found by measuring the contact angle of liquid polystyrene on coating substrates. The values are given in Table 10.
| Refractory Filler | Contact Angle of Liquid Polystyrene (°) |
|---|---|
| Alumina (bauxite) | 82 |
| Quartz powder | 65 |
| Talc powder | 53 |
| Graphite powder | 25 |
The lower the contact angle, the better the wettability and the easier the liquid decomposition products are drawn into the coating and transported out. I found that adding a small amount of graphite and a surfactant to the coating greatly reduced the contact angle and improved the escape of liquid residues, without significantly harming the coating strength.
The principle of decomposition-product transport through the coating can be described by Darcy’s law. The volume flow rate Q of a liquid through a porous coating is given by
$$Q = \frac{k \cdot A \cdot \Delta P}{\eta \cdot L}$$
where k is the permeability of the coating, A is the effective area, ΔP is the pressure gradient, η is the viscosity of the liquid, and L is the coating thickness. Lower viscosity and higher permeability naturally increase the escape rate. The pressure gradient is provided by the vacuum negative pressure, which I will discuss next.
Effect of Sand Size and Ventilation
I compared three sand size ranges under identical conditions. Table 11 shows the results.
| Sand Mesh | Centrality (%) | Filling Time (s) | Back-spraying | Carbon Defect Rate (%) |
|---|---|---|---|---|
| 20–30 | 85 | 68 | No | 6 |
| 30–40 | 82 | 78 | No | 10 |
| 40–50 | 80 | 102 | Severe | 22 |
Coarser sand provides higher permeability for gases and liquids. At the same time, coarse sand also influences the heat transfer and the pressure distribution in the mold. In lost foam casting, the vacuum is applied to the whole sand mass, so the local pressure around the decomposing pattern is strongly influenced by the sand resistance. Fine sand restricts the flow of decomposition products, creating a higher local back-pressure that slows the metal front and forces more liquid carbon residue into the metal.
The permeability of a sand compact can be approximated by the Kozeny-Carman equation:
$$K = \frac{\varepsilon^3}{C \cdot (1-\varepsilon)^2} \cdot \frac{d^2}{36}$$
where ε is the porosity, C is a constant, and d is the average grain diameter. Because permeability scales with the square of the grain diameter, small changes in sand fineness cause large changes in venting capacity. In practice, I recommend using a coarser sand together with effective vacuum extraction. A sand size of 20–30 mesh was found to give the best balance between defect minimization and surface finish.
Another important observation is the effect of sand recycling. After several cycles of use, dry sand accumulates fine dust, ash, and condensed decomposition products. The ash content and loss-on-ignition of the sand increase steadily. I measured the ash content of new sand and a sand that had been in service for six months: it increased from 0.35% to 1.85%, and the loss-on-ignition increased from 0.20% to 1.30%. This reduces the permeability of the sand bed and increases the tendency for carbon defects. Therefore, regular sand regeneration or periodic replacement is necessary, typically every six months in our production environment.
Effect of Vacuum Pressure
The vacuum pressure in the lost foam casting mold is a key parameter that directly drives the removal of decomposition products. I tested different vacuum settings from 0 to 0.08 MPa while maintaining all other conditions constant. The results in Table 12 show that negative pressure strongly affects defect formation and also influences sand stability and metal penetration.
| Vacuum Pressure (MPa) | Carbon Defect Condition | Other Defects |
|---|---|---|
| 0 | Severe carbon defects, incomplete filling | Mold collapse |
| 0.02 | Carbon defects severe | Metal splash, mold collapse |
| 0.035 | Surface carbon defects | No collapse |
| 0.05 | No surface defects | Slight sand penetration |
| 0.06 | No defects | None |
| 0.07 | No defects, but internal carbon spots | Sand fusion near thin sections |
| 0.08 | No surface defects | Severe sand penetration, internal carbon defects |
Too low a vacuum does not provide enough pressure difference to evacuate the pyrolysis gases and liquids, so they remain in the mold and cause defects. Too high a vacuum creates a strong “wall effect,” where the metal is pulled preferentially along the sand-wall interface, trapping some decomposition products in the interior of the casting. The optimal vacuum window for my castings was found to be 0.05–0.06 MPa. For thick-section castings, a slightly lower vacuum can be used; for thin-wall castings, a slightly higher vacuum may improve filling, but careful attention must be paid to avoid internal carbon defects.
The gas flow through the sand under a negative pressure can be modeled by Darcy’s law for compressible gas:
$$Q_g = \frac{K_g \cdot A \cdot (P_1^2 – P_2^2)}{2 \cdot \eta_g \cdot L \cdot P_0}$$
where Qg is the gas volume flow rate, Kg is the gas permeability of the sand, P1 and P2 are the pressures at the coating surface and the vacuum side, P0 is the reference pressure, ηg is the gas viscosity, and L is the sand bed thickness. This equation emphasizes the need for a sufficient pressure difference and a permeable sand path.
Effect of Gating System and Casting Geometry
The design of the gating system in lost foam casting has a direct influence on where the carbon-rich decomposition products end up. A top gate system allows the metal to fill from the bottom upward, pushing decomposition products ahead of the front and out through the top gates. This is beneficial for surface carbon defects but may cause turbulent flow and internal entrainment. A bottom gate system promotes laminar filling, but the last-filled region, which is at the top, is the region where the metal is coolest and where all the residual liquid products accumulate. In my tests with a large transmission housing, a bottom gate resulted in severe carbon defects at the top mounting flanges. Changing to a top gate eliminated those defects, although it slightly increased the occurrence of entrained inclusions in the lower parts. The best solution was a multi-step gating system combining top gates for the critical areas and a bottom runner for uniform rising of the metal.
The geometry of the casting also matters. I introduced a “modulus” parameter, defined as
$$M = \frac{V}{S}$$
where V is the volume of the casting and S is the external surface area. A casting with a smaller modulus has a relatively larger surface area through which decomposition products can escape. Table 13 compares two different castings produced in the same workshop.
| Casting | Volume (cm³) | Surface Area (cm²) | Modulus M (cm) | Scrap Rate due to Carbon Defects (%) |
|---|---|---|---|---|
| Transmission housing | 14500 | 9600 | 1.51 | 12 |
| Pump body | 6200 | 5200 | 1.19 | 6 |
As expected, the casting with the smaller modulus had a lower tendency to form carbon defects because the larger relative surface area provided more exit paths for the decomposition products.
Formation Mechanism of Carbon Defects
Based on all these experiments and analyses, I have formulated a comprehensive mechanism of carbon defect formation in lost foam casting iron castings. When molten iron at 1400–1500°C contacts the foam pattern, the polymer (EPS, PMMA, or copolymer) undergoes rapid thermal decomposition. The decomposition creates a transient gap between the metal and the remaining foam. The products are a mixture of gas, liquid tar, and solid carbon. Under ideal conditions, the gases escape through the porous coating and the sand layer, while the liquids also wet the coating and are drawn out by capillary action and the vacuum pressure. However, if the metal advance rate is faster than the escape rate, the liquid residues accumulate at the metal front. Because the metal is in contact with the unvaporized foam, the local atmosphere is highly reducing and deficient in oxygen. The liquid polymer undergoes secondary pyrolysis and cyclization, forming polycyclic aromatic hydrocarbons and eventually solid carbon. This carbon-rich sludge is pushed ahead of the metal until it reaches a dead end, such as the last-filled region, a corner, or a junction with a riser. There, it solidifies along with the metal, creating the black carbon inclusions.
Mathematically, the condition for avoiding carbon defects can be expressed as a balance between the rate of decomposition product generation, G, and the escape rate, E:
$$G = \rho_f \cdot A_f \cdot v_f \cdot C_{vol}$$
where ρf is the foam density, Af is the area of the decomposition front, vf is the linear filling velocity of the metal, and Cvol is the mass of volatile products per unit volume of foam. The escape rate E is controlled by the coating and sand permeability, the pressure differential, and the wetting properties. If E ≥ G, the casting is sound. If E < G, carbon defects are likely. This is why a combination of low-density foam, low-carbon polymer, high pouring temperature, high coating permeability, high wetting ability, coarse sand, and optimal vacuum is required.
Prevention Measures and Industrial Practice
I have implemented the following practical measures in my foundry to reduce and eliminate carbon defects in lost foam casting iron castings.
Selection of Low-Carbon Copolymer and Low Pattern Density
I switched from pure polystyrene to a copolymer containing approximately 70% EPS and 30% PMMA. This reduces the carbon content from 92% to about 77%, and introduces oxygen atoms that promote cleaner decomposition. At the same time, I optimized the pre-expansion process to achieve a pattern density of 0.020–0.022 g/cm³ for the transmission housing. This is the lowest density that still maintains sufficient mechanical strength for handling and coating. The pre-expansion parameters for the copolymer are given in Table 14.
| Parameter | Value |
|---|---|
| Bead brand | Mitsubishi Chemical copolymer (JM310) |
| Average bead diameter | 0.8–1.0 mm |
| Pre-expansion ratio | 18–25× |
| Pre-expansion steam pressure | 0.15–0.25 MPa |
| Pre-foam density | 0.018–0.022 g/cm³ |
| Storage time after pre-expansion | 2–4 days |
| Drying temperature | 40–50°C |
| Drying time | 6–8 h |
High Pouring Temperature
I raised the pouring temperature for iron castings to 1480–1520°C, which is about 30–50°C higher than conventional sand casting. This higher temperature ensures that the foam pattern is fully gasified, leaving very little liquid residue. However, I must carefully monitor the coating quality to avoid metal penetration and sand fusion at such high temperatures.
Optimized Coating and Sand System
I developed a coating with a high-temperature permeability greater than 0.30 cm⁴/g·min and with improved wettability to liquid polystyrene. This was achieved by adding 8–10% pyrolitic graphite and a surfactant to the alumina-based coating. The coating thickness was carefully controlled at 1.5–2.0 mm. For the sand, I use a 20–30 mesh quartz sand with a vibration frequency and amplitude adjusted to achieve a density of 1.55–1.65 g/cm³. I also installed a sand regeneration system that removes fines and dust by an air classifier, and I replace about 20% of the sand with fresh sand every month to maintain consistent permeability.
Appropriate Vacuum Profile
The vacuum pressure was optimized to 0.055 MPa for the box-shaped castings. I also introduced a two-stage vacuum profile: a high vacuum of 0.06 MPa during the first half of filling to quickly evacuate the initial burst of pyrolysis gases, followed by a reduced vacuum of 0.04 MPa during the second half to avoid excessive metal front deformation and the wall effect. This profile significantly reduced both surface and internal carbon defects.
Gating and Riser Design
I use a ceramic foam filter in the gating system to capture solid carbon particles carried by the metal. For castings with large top surfaces, I add small breakout risers at the highest points. These risers act as carbon traps: the liquid residues and carbon-rich slag float up into the riser, keeping the main casting clean. On the transmission housing, I added four cylindrical risers of 50 mm diameter and 80 mm height at the corners of the top flange. Visual inspection and machining confirmed that these risers collected almost all the carbon contaminants, and the casting remained free of defects.
Application in Production
After implementing these measures, the scrap rate of the transmission housing due to carbon defects dropped from 38% to less than 2%. The overall casting yield, surface quality, and machining performance improved dramatically. Similar results were achieved on other castings, including pump bodies and motor housings. The lost foam casting line now operates efficiently, producing high-quality iron castings that meet international standards.
Summary and Conclusions
Through my extensive investigation, I have reached the following conclusions regarding carbon defects in lost foam casting iron castings:
1. Carbon defects are carbon-rich inclusions formed by the thermal decomposition products of the foam pattern that fail to escape the coating and are later solidified in the last-filled regions of the casting. They contain up to 84% carbon and no oxide species, confirming their origin from the polymeric pattern.
2. Carbon defects seriously compromise the mechanical properties, hardness, pressure tightness, and machinability of lost foam casting iron castings. In my experiments, tensile strength was lowered by as much as 32%, and hardness by 17%.
3. The primary material factor is the chemical composition of the pattern. High-carbon EPS produces far more liquid and solid residues than lower-carbon copolymers of EPS and PMMA. The pattern density also plays an additive role: the more foam material per unit volume, the greater the amount of decomposition products that must be removed.
4. Pouring temperature is a highly sensitive parameter. Higher temperatures shift the decomposition toward volatile gaseous products, reducing liquid and solid residues. The critical temperature for the studied castings was about 1440°C, with 1480°C recommended for safe production.
5. Coatings and sand are the escape channels for decomposition products. High high-temperature permeability and good wettability of the coating by liquid polymer greatly increase the escape rate. Coarse sand with high permeability also helps. The vacuum pressure must be optimized to drive the products out without creating adverse wall effects. A value of 0.05–0.06 MPa was found suitable for medium-sized iron castings.
6. Proper gating design, the addition of carbon-trap risers, controlled casting geometry, and regular sand maintenance are essential practical measures to suppress carbon defects.
I acknowledge that my research has limitations. The theoretical analysis of the complex thermal-physical-chemical interactions is still incomplete. The domestic raw materials in my foundry, although similar in nominal properties to imported ones, behaved worse in practice, pointing to a need for deeper molecular-level investigation and standardization. In the future, I plan to establish a dedicated laboratory for testing foam beads and coatings, and to develop additives that can modify the decomposition products to be more easily removable. This will further advance the science and application of lost foam casting technology.
In summary, lost foam casting remains one of the most promising casting methodologies for the 21st century. Solving the carbon defect problem is the master key to unlocking its full potential, and I am confident that the findings presented here will help many foundries achieve high-quality, low-cost castings using lost foam casting.