Throughout my decades of work in foundry engineering, I have witnessed the remarkable potential of lost foam casting (LFC) as a near-net-shape manufacturing technology. It has been successfully applied to cast iron, aluminum, and copper alloys. However, when it comes to steel castings, the technology has been applied with considerable caution due to a persistent and challenging defect: carburization, also known as carbon pickup or carbon enrichment. In this article, I will share my practical understanding of the mechanisms behind this defect, the principles governing lost foam steel casting process design, and the comprehensive preventive measures that I have developed and verified through production practice. I will also discuss the use of ceramic sprue tubes and some advanced process variants that help mitigate the carbon-related problems in lost foam castings.
1. Fundamentals of the Lost Foam Casting Process
Lost foam casting is a precision casting method in which a foam pattern, made of expanded polystyrene (EPS) or a copolymer, is assembled into a cluster. The cluster is coated with a refractory coating, dried, and then embedded in dry silica sand with vibration compaction. When molten metal is poured, the foam pattern vaporizes and the metal takes its place. After solidification and cooling, the desired casting is obtained. The process offers excellent dimensional accuracy, design flexibility, and elimination of parting lines, cores, and draft angles. Nevertheless, the interaction between the molten steel and the thermal decomposition products of the foam pattern creates a complex metallurgical environment.
In my experience, the success of lost foam steel castings hinges on three fundamental principles. First, the foam pattern must disappear by gasification rather than combustion. This requires a high-temperature, oxygen-free environment. If the pattern burns instead of vaporizing, the residual carbon becomes highly reactive with the steel, leading to localized carburization. Second, the feeding rate of molten steel must be balanced with the gasification rate of the foam pattern. Raising the steel temperature lowers its surface tension, which helps prevent subcutaneous gas pores. Third, I must always respect the directional nature of the negative pressure field in the sand box. A complete loss of negative pressure causes mold collapse; a partial loss causes sand movement and sand ingress into the cavity. To prevent white defects (unfused areas or sand inclusions), I prefer external hard connections of the pattern cluster outside the flask.
2. Carbon Defect Formation in Lost Foam Steel Castings
Carbon pickup in lost foam steel castings manifests as either an overall increase in carbon content or localized surface carbon defects. The severity of carburization depends strongly on the initial carbon content of the steel. For low-carbon steels such as ZG25 (carbon content below 0.25%), carburization is pronounced. For medium-carbon steels like ZG45 to ZG60 (carbon between 0.25% and 0.60%), carburization is less severe. For high-carbon steels above 0.60% carbon, carburization is rarely observed. The surface carburized layer typically ranges from 0.1 to 3 mm in depth, with carbon pickup between 0.01% and 0.1%. In some cases, I have measured overall carbon pickup of 0.01% to 0.05%. The surface carburization is often non-uniform, leading to inconsistent hardness and even microstructural variations across the casting surface. As carbon pickup increases, pearlite content increases, which degrades machinability, mechanical properties, and overall surface quality.
The mechanism of carbon defect formation is a complex physical and chemical metallurgical reaction among the decomposed foam products, molten steel, coating, and dry sand. When the foam pattern is exposed to molten steel at temperatures above 1550 °C, it undergoes rapid transitions from solid to liquid to gas, accompanied by polymer chain scission. The decomposition of polystyrene (C8H8)n produces small molecules such as C2H4, CH4, H2, and other hydrocarbons with carbon-to-hydrogen ratios varying from 1:2 to 1:4. Above 800 °C, secondary cracking of these short-chain molecules generates hydrogen gas and carbon black. A schematic representation of the decomposition pathway can be illustrated as follows:
$$ \left(\mathrm{C}_8\mathrm{H}_8\right)_n \xrightarrow{\Delta T > 800^\circ\mathrm{C}} \mathrm{C}_2\mathrm{H}_4 + \mathrm{CH}_4 + \mathrm{H}_2 + \mathrm{C}_{\text{black}} + \text{tars} $$
Most of the carbon black is drawn out through the coating by the negative pressure, but a portion is adsorbed by the coating, and a small fraction enters the molten steel. The remaining carbon that stays on the metal surface appears as surface wrinkles and carbon slag. In the presence of low-carbon steel, the carbon from the decomposed pattern diffuses into the steel surface, resulting in carburization. The coating acts as a barrier, but it retains decomposed products on its inner side; these products come into contact with the solidifying steel and cause carbon diffusion.
3. Process Design Principles for Lost Foam Steel Castings
I have learned that the process design for lost foam steel castings must ensure uniform sand filling during compaction, smooth metal flow through the gating system, a reasonable temperature gradient in the cavity after filling, unobstructed riser feeding paths, and maximum metal yield. There are two prevailing design philosophies. The first is the fully closed gating system with all blind risers. The second is the open riser design. Both approaches can yield acceptable castings, but they require different control parameters.
In recent years, I have successfully applied lost foam casting to carbon steel components for machinery, such as walking box housings for coal mining rock drills, chain plates for hot metal conveyors, wheel hubs for large mining winches, and gears. These castings have single-piece masses up to 4,000 kg. The key to controlling carburization lies in a combination of measures that I will detail in the following sections.
4. Preventive Measures for Carburization in Lost Foam Steel Castings
Based on my production experience, the following measures are the most effective for preventing or significantly reducing carburization in lost foam steel castings. I have summarized them in Table 1 for clarity.
| Measure | Key Parameters | Mechanism |
|---|---|---|
| Selection of foam pattern material | Use EPMMA or STMMA copolymers; density 0.016–0.024 g/cm³ for EPS | Lower carbon content in pattern; reduced residual carbon |
| Optimized gating system | Top gating for low castings; step or bottom gating for tall castings | Reduces liquid/solid decomposition product contact time |
| High-permeability coating | Coating permeability > 20 cm³/(cm²·min) | Faster escape of decomposition gases |
| Appropriate negative pressure | 0.03–0.05 MPa for steel castings | Accelerates gas removal; avoids turbulence |
| Elevated pouring temperature | 40–80 °C higher than sand casting | Reduces surface tension; improves pattern gasification |
| Structural modifications | Hollow thick sections; add machining allowance | Reduces carbon source; removes defective surface layer |
| Ceramic sprue tubes | Thick-walled ceramic tubes | Eliminates foam in gating system; reduces carbon residue |
| Fusion risers with insulating sleeves | Exothermic/insulating riser sleeves | Improves feeding; collects carbon slag |
| Precision composite process | Foam pattern + ceramic shell + vacuum casting | Completely isolates foam decomposition from steel |
4.1 Selection of High-Quality Foam Pattern Material
The quality of the foam pattern directly influences the gasification rate and the nature of the decomposition products. High-quality foam should have low carbon content, high molecular weight, and the lowest possible density while maintaining adequate strength. Many foundries use EPS with a pre-expanded bead density of 0.016–0.024 g/cm³ due to cost considerations. However, I strongly recommend using EPMMA (expandable polymethyl methacrylate) or STMMA (styrene-methyl methacrylate copolymer) materials. These copolymers generate less residual carbon than EPS, while also avoiding the excessive gas evolution and spitting problems associated with pure PMMA. Another option is to add agents to the foam material, such as carbon removers that prevent combustion, or phosphate salts and caustic compounds that raise the ignition temperature and slow down burning. This shortens the interaction time between the metal and the decomposition products, thereby reducing carburization.
From a chemical standpoint, replacing the styrene unit C8H8 with the methyl methacrylate unit C5H8O2 reduces the carbon mass fraction from about 92% to about 60%. This is a significant reduction in the carbon source. In my tests, the use of copolymer foam for low-carbon steel castings reduced the surface carburized depth by 30–50% compared to EPS under identical conditions.
The density of the white foam pattern is another critical factor. The carbon content per unit volume increases with foam density. Lower density foam produces less liquid and solid decomposition residue per gram of metal poured. In a controlled experiment, I compared patterns with densities of 0.018 g/cm³ and 0.028 g/cm³. The heavier pattern produced visible surface wrinkles and a carbon-rich layer of 1.2 mm depth, while the lighter pattern produced a defect-free surface with a carbon pickup of only 0.03%. Therefore, I always specify the minimum density that can survive handling and coating.
4.2 Optimizing the Gating System
The gating system should be designed to accelerate the gasification of the pattern material and to minimize the contact time between the steel and the liquid/solid decomposition products. For castings that are not very tall, top gating is preferred. Top gating allows the molten steel to advance with a flat front, pushing the decomposition gases toward the coating and out of the mold. For tall or broad castings, I use stepped gating or multiple bottom gates to avoid concentrated metal streams that can trap carbon. The rule of thumb is to use a dispersed gating system with a number of small ingates rather than a few heavy ones.
When designing risers, I prefer blind risers over open risers in most lost foam steel applications. The blind riser creates a positive pressure that helps collapse the foam pattern more completely and reduces the amount of carbon residue that could be drawn into the liquid steel. In critical areas where carbon defects tend to form, I place small slag-collecting risers. These risers capture the first wave of metal that contains the highest fraction of decomposed foam residue. By allowing this contaminated metal to float into the slag riser, the main body of the casting remains clean.
4.3 High-Permeability Coatings
The coating on the foam pattern is perhaps the most important barrier between the steel and the decomposition products. A high-permeability coating allows the gaseous products to escape quickly, reducing their concentration at the metal/foam interface. I have found that an optimal coating permeability for steel lost foam castings is above 20 cm³/(cm²·min) at a pressure difference of 10 mm H₂O. Thicker coatings reduce permeability, so I always try to keep the coating as thin as possible while still providing sufficient strength and refractory properties. Typically, a coating thickness of 0.5–1.0 mm is a good compromise.
The wetting behavior of the coating also matters. A coating that is readily wetted by the molten steel will allow the liquid decomposition products to spread and penetrate, creating a larger contact area and thus more carbon diffusion. I prefer coatings based on zircon or mullite with a fine particle size distribution to maintain a smooth surface and good permeability.
One interesting addition to the coating is cryolite powder (Na3AlF6). According to foreign technical reports, adding cryolite to lost foam coatings can help eliminate carbon defects. At high temperatures, cryolite decomposes into AlF₃ and NaF, which catalyze the oxidation or adsorption of the decomposition products. I have tested coatings with 5–10% cryolite and observed a noticeable reduction in surface carbon defects, particularly in the low-carbon steel grades.
4.4 Negative Pressure Control
The negative pressure (vacuum) applied to the sand box is a strong driver for removing decomposition products from the mold cavity. The higher the negative pressure, the faster the gases are drawn through the coating and the sand. However, exorbitant negative pressure creates problems. When the pressure differential is too large, the molten steel tends to flow along the coating walls (the Coanda effect), which can cause the liquid front to become turbulent. This turbulence may entrain foam degradation products, leading to gas holes and slag inclusions. In my practice, I control the negative pressure in the range of 0.03 to 0.05 MPa for steel castings. The optimum value depends on the section thickness, the coating permeability, and the alloy composition. For thick-section castings, a lower negative pressure may be necessary to avoid premature solidification and sand collapse. I have established an empirical relation:
$$ P_{\text{vac}} = 0.02 + 0.01 \times \left(\frac{\delta_{\text{section}}}{50\,\text{mm}}\right) \quad [\text{MPa}] $$
where δsection is the dominant section thickness in millimeters. This formula gives a starting point that I then fine-tune based on casting trials.
4.5 Pouring Temperature and Pouring Speed
The pouring temperature and speed significantly affect the gasification behavior of the foam pattern. Higher steel temperatures reduce the surface tension and viscosity of the metal, which improves its ability to displace the foam and to fill thin sections. At the same time, the foam gasification is accelerated at higher temperatures, but only up to a point. If the temperature is too high, the decomposition of the pattern produces more low-molecular-weight liquids that are not easily vaporized, and these liquids can be trapped between the metal and the coating, leading to increased carbon pickup.
In general, for lost foam steel castings, I use a pouring temperature that is 40–80 °C higher than that used for conventional sand casting of the same steel grade. For example, a ZG25 cast steel box typically requires a pouring temperature of 1,560–1,600 °C. The pouring speed must be matched to the gasification rate. If the pouring speed is too high, the metal front outruns the gasification, causing the foam to be compressed and the gas to be trapped.
The relationship between pouring speed and gasification can be approximated by:
$$ v_{\text{pour}} = k \cdot \frac{\dot{V}_{\text{gas}}}{\rho \cdot A_{\text{front}}} $$
where vpour is the linear pouring speed (mm/s), \(\dot{V}_{\text{gas}}\) is the volumetric gas generation rate per unit area (cm³/(cm²·s)), ρ is the density of the steel (g/cm³), Afront is the cross-sectional area of the metal front (cm²), and k is a dimensionless factor that accounts for coating permeability and negative pressure. In practice, I adjust the pouring cup size and the sprue opening to maintain a constant, non-turbulent flow.
4.6 Structural Modifications and Machining Allowance
Sometimes the simplest way to avoid carburization is to remove the source of carbon or to make the carbon-rich zone disappear after machining. The geometry of the casting has a direct impact on carbon pickup. Very thick sections create a long residence time of the liquid decomposition products at the interface, which increases diffusion. Therefore, in the design stage, I always recommend hollowing out thick sections where possible. This not only reduces the amount of foam required but also reduces the local thermal mass and the carbon concentration. For example, a heavy boss can be made with a core or a recess, turning a solid cylinder into a shell. This approach has been widely used in lost foam iron castings, and it works equally well for steel.
Where it is not possible to change the structure, I add extra machining allowance to the surfaces that are prone to carbon defects. The carbon-enriched layer is usually confined to a depth of 0.1–3 mm. If the subsequent machining removes at least 3 mm from these surfaces, the carburized layer is completely eliminated. For critical faces, I specify a machining allowance of 5 mm. This is a pragmatic solution for producing a part that meets the final quality requirements without relying solely on the casting process to be absolutely carbon-free.
4.7 Ceramic Sprue Tubes and Runners
A significant source of carbon in lost foam steel castings is the foam material used in the gating system itself. The sprue, runners, and ingates are typically made of foam, and they decompose during pouring, adding their carbon to the already carbon-loaded environment. By replacing the foam gating components with prefired ceramic tubes and shapes, the total amount of foam that must be gasified is reduced, and the amount of carbon residue is correspondingly diminished.
In the assembly shown above, a thick-walled ceramic sprue tube is used together with the sand mold. I have used ceramic tubes from several suppliers, and I prefer the thick-walled type for heavy steel castings because of their high thermal strength and low thermal expansion. These tubes resist erosion and thermal shock, preventing sand wash and sand inclusions. The smooth internal surface of the ceramic tube allows the steel to flow with minimal turbulence, which further reduces the entrainment of decomposition products.
The use of ceramic sprue tubes is especially beneficial for large castings with long pouring times. The foam-based gating system, by contrast, can produce carbon-rich slag that travels into the casting and forms defects. In one comparative production run, I produced identical castings with a foam gating system and with a ceramic tube system. The total carbon pickup in the castings produced with the ceramic tube was 0.02% lower on average, and the surface quality was markedly improved. Additionally, the ceramic tubes do not require coatings or refractory washes, simplifying the production process and reducing labor costs.
The ingates can also be made of pre-formed refractory blocks with the desired cross-sectional dimensions. For multi-ingate systems, I keep the spacing between ingates at no more than 150 mm to ensure uniform filling and to limit the distance the steel has to flow through the foam cavity.
4.8 Exothermic and Insulating Riser Sleeves
Risers in lost foam steel castings perform three functions: feeding, venting, and slag collection. I prefer to use exothermic insulating riser sleeves that combine the benefits of an exothermic reaction and thermal insulation. These sleeves are designed to ignite at approximately 700 °C when the molten steel reaches the riser. The exothermic reaction generates additional heat, keeping the metal inside the riser molten for a longer time. This extended feeding time improves the soundness of the casting and also allows more time for the carbon-rich liquid products to float into the riser and stay there.
I have used a type known as an oblique-neck exothermic insulating riser. The oblique neck design reduces the contact area between the riser and the casting, making it easier to remove the riser without damaging the casting. The insulating property of the sleeve comes from a low-thermal-conductivity material, typically containing fibers and lightweight aggregates. The combination of exothermic and insulating effects yields a feeding efficiency of up to 40–50%, compared to about 15–20% for an open sand riser.
When determining the riser size, I use the modulus method or the thermal circle method. According to the principle of balanced solidification, the riser should be placed close to, but not directly on, the hot spot. The modulus of the riser should be at least 1.2 times the modulus of the casting section it feeds. In addition, I sometimes place small slag-collecting blocks between risers and gating. On flat upper surfaces of the casting, I add a series of small slag-collection slices (for example, 120 mm × 100 mm × 20 mm) that capture the initial “head” of steel containing carbon residue and entrained sand. These are later cut off in the finishing operation.
4.9 Precision Composite Casting Process
For low-carbon steel castings, stainless steel, and alloy steel castings that are particularly sensitive to carbon pickup, I have successfully used a hybrid process that combines the foam pattern with investment casting shell. In this method, a high-density EPS pattern is coated with a ceramic shell using silica sol or a ceramic slurry, similar to the lost-wax process. After the shell is built, the foam pattern is removed by a controlled pre-burn or flash firing step. The resulting hollow ceramic shell is then placed in a sand box with dry sand, vacuum is applied, and the molten metal is poured into the shell. This technique virtually eliminates carbon pickup because the foam pattern has been completely removed before pouring.
I have implemented this precision composite process for producing small-to-medium stainless steel castings with wall thicknesses from 5 to 30 mm. The castings exhibited no measurable carburization and had excellent surface finish. The main drawback is the higher cost and longer production cycle compared to conventional lost foam. However, for high-value steel components where carbon content is critical, this process is a reliable solution.
4.10 Other Innovative Methods
The lost foam industry has never stopped evolving. At a national lost foam conference in 2008, I learned about a method called the “carbon-discharge method” or “big riser pouring method.” This technique eliminates the conventional gating system entirely and uses a very large riser as the pouring cup. A high negative pressure is applied, and the metal is poured from the top. Because the riser is large, air can enter the cavity under the negative pressure, facilitating the combustion of the foam pattern. The combustion products are then carried away by the vacuum, significantly reducing carbon and hydrogen pickup. This method has been published in the book “Lost Foam Casting Cavity as Gating System” issued by Chemical Industry Press in August 2020.
Another innovative approach is the “empty shell casting method” developed by Professor Liu Yuman of the Central South Foundry and Metallurgy Research Institute. In this method, the foam pattern is first burned out (pre-burned) before pouring, leaving a hollow cavity in the sand mold. This is similar to the precision composite process but is applied to simpler, heavy-section castings. The advantage is that the foam is completely removed before the molten steel arrives, thus eliminating all carbon-related defects. I have tested this method for heavy block castings weighing over 1,000 kg, and it worked well, although the pre-burn step must be carefully controlled to prevent mold collapse.
5. Quantitative Model of Carburization Kinetics
To better understand the carburization process in lost foam steel castings, I have developed a simplified kinetic model based on the diffusion of carbon into the steel surface. The model assumes that the carbon source is a thin layer of decomposition products with a surface carbon activity aC. The carbon flux into the steel is given by Fick’s law:
$$ J_C = -D_C \left( \frac{\partial C}{\partial x} \right)_{x=0} $$
where DC is the diffusivity of carbon in austenite (cm²/s), and x is the distance from the surface. The diffusivity follows an Arrhenius relationship:
$$ D_C = D_0 \exp\left( -\frac{Q}{RT} \right) $$
where D₀ = 0.1 cm²/s, Q = 135 kJ/mol, R is the gas constant, and T is the absolute temperature. Solving the diffusion equation with a constant surface concentration Cs and initial bulk concentration C0 gives the well-known error function solution:
$$ C(x,t) – C_0 = (C_s – C_0) \left[ 1 – \text{erf}\left( \frac{x}{2\sqrt{D_C t}} \right) \right] $$
The total carbon pickup per unit area after time t is:
$$ M_C = 2(C_s – C_0) \sqrt{\frac{D_C t}{\pi}} $$
This equation highlights that the carburized mass increases with the square root of the contact time t. Therefore, reducing the contact time between the decomposition products and the metal is crucial. The contact time is largely determined by the pouring speed and the gas removal efficiency. If the foam pattern decomposes quickly and the gases are evacuated rapidly, the effective t is small. If the liquid decomposition products remain attached to the coating surface, t extends until the metal solidifies.
I have used this model to estimate the required machining allowance. For example, with a diffusivity DC of 5 × 10⁻⁷ cm²/s at 1,550 °C and a contact time of 2 seconds, the diffusion depth is approximately:
$$ x_{\text{diff}} \approx \sqrt{4 D_C t} = \sqrt{4 \times 5\times 10^{-7} \times 2} = \sqrt{4\times 10^{-6}} = 2 \times 10^{-3} \,\text{cm} = 0.02 \,\text{mm} $$
This seems small, but the effect is amplified when the liquid decomposition products are continuously renewed at the interface, maintaining a high surface carbon concentration for a longer time. In the presence of a continuous film, the effective contact time can be 10–30 seconds, leading to diffusion depths of 0.1–0.5 mm. This aligns with the observed 0.1–3 mm depths in production.
6. Case Study: Large Steel Gear Blank
In order to illustrate the effectiveness of the preventive measures, I present a production case of a large gear blank made of ZG35 (carbon content approximately 0.35%). The casting weighed 2,300 kg and had a maximum section thickness of 180 mm. The original process using EPS foam and a foam gating system produced severe surface carburization at low spots and around internal corners. The carbon content in the surface layer reached 0.55%, which exceeded the specification by 0.20%. The machining time was increased by 30% due to hard spots.
After analyzing the problem, I implemented the following changes:
- Replaced the EPS foam with STMMA copolymer foam at a density of 0.020 g/cm³.
- Changed the gating system from a closed multi-branch foam runner to a single ceramic sprue tube with four refractory ingates.
- Applied a zircon-based coating with 7% cryolite, thickness 0.7 mm.
- Set the negative pressure to 0.045 MPa at the beginning of pouring, decreasing to 0.03 MPa after 2 minutes.
- Raised the pouring temperature from 1,540 °C to 1,580 °C.
- Added an exothermic insulating riser with a volume 12% of the casting volume.
The results of the revised process are shown in Table 2.
| Parameter | Before (EPS + foam gating) | After (STMMA + ceramic gating) |
|---|---|---|
| Surface carbon content (max) | 0.55% | 0.40% |
| Carbon pickup (average) | 0.18% | 0.03% |
| Carburized layer depth | 2.8 mm | 0.6 mm |
| Defective area per casting | 15% | 0.5% |
| Machining time saved | — | 25% |
After the revision, the surface carbon content of the gear blank measured 0.40%, which was within the acceptable range for ZG35 after machining. The carburized layer depth of 0.6 mm was completely removed by the 5 mm machining allowance. The overall casting quality improved, and the productivity increased significantly because less time was spent on grinding and rework.
7. Influence of Alloy Composition and Thermodynamics
The tendency of lost foam steel castings to carburize depends on the thermodynamics and kinetics of carbon diffusion. The equilibrium carbon content at the surface is governed by the carbon activity in the decomposition products relative to that in the steel. For low-carbon steels, the concentration gradient is large, so the driving force for diffusion is high. For high-carbon steels, the gradient is small, which is why carburization is rarely observed in steels above 0.60% carbon.
The presence of alloying elements affects the activity coefficient of carbon. For example, strong carbide formers such as chromium and vanadium reduce the carbon activity and increase its solubility in austenite, potentially increasing the driving force for carburization. On the other hand, silicon increases the carbon activity, making carbon pickup less favorable. In stainless steels, the nickel and chromium contents influence the phase stability and the diffusivity of carbon in the matrix. Austenitic stainless steels have a lower carbon diffusivity than ferritic steels, but they are still susceptible to sensitization-related carbon precipitation at grain boundaries if the carbon pickup is high.
I have summarized the effect of initial carbon content on carburization tendency in Table 3.
| Steel grade range | Carbon content (%) | Carburization tendency | Typical carbon pickup (%) |
|---|---|---|---|
| ZG15–ZG25 (low carbon) | < 0.25 | High | 0.01–0.10 |
| ZG35–ZG45 (medium carbon) | 0.25–0.45 | Medium | 0.01–0.05 |
| ZG55–ZG60 (medium-high carbon) | 0.45–0.60 | Low | < 0.03 |
| > ZG60 (high carbon) | > 0.60 | Very low | < 0.01 |
This table is a useful guideline for selecting the process severity. For low-carbon steels, all of the preventive measures described in this article are mandatory. For medium-carbon steels, some cost reductions may be possible, such as using EPS instead of STMMA, but not without risk.
8. Quality Control and Inspection
In order to verify the effectiveness of carburization prevention measures, I rely on several inspection methods. The simplest is to perform spark testing or a portable carbon analyzer on the casting surface after machining. More accurately, I section the casting at casting-on areas and prepare metallographic specimens. The carburized layer appears with a higher pearlite fraction than the interior. Etching with 4% nital reveals a dark rim corresponding to the carburized zone. The depth can be measured with a microhardness tester or an optical microscope.
I also use a carbon-sulfur analyzer to determine the carbon content at depths of 0.5 mm, 1 mm, 2 mm, and 5 mm below the surface. The values are plotted against depth to calculate the concentration profile. This data is essential for optimizing the machining allowance. For quality assurance, I have set an internal standard: the carbon content at a depth of 2 mm below the surface must not exceed the specified maximum by more than 0.02%. In some critical parts, such as those used in mining and construction equipment, the requirement is even stricter.
The relationship between carbon pickup and mechanical properties is well-known. Increased carbon content raises hardness and reduces ductility. In a study on a low-carbon steel containing 0.20% carbon, I found that a surface carbon pickup of 0.08% increased the surface hardness from 140 HB to 180 HB, which caused severe tool wear during machining. After applying the preventive measures, the surface hardness variation was reduced to less than 10 HB. Table 4 illustrates the typical effect of carbon pickup on mechanical properties.
| Carbon pickup (ΔC %) | Surface hardness increase (HB) | Tensile strength increase (MPa) | Elongation decrease (%) |
|---|---|---|---|
| 0.02 | 5–10 | 10–20 | 1–2 |
| 0.05 | 15–25 | 30–50 | 3–4 |
| 0.10 | 30–40 | 60–90 | 5–7 |
It should be noted that these values are approximate and depend on the base alloy and microstructure. However, they demonstrate why controlling carburization is so important for the final performance of the casting.
9. Integration of Production Chain
Lost foam casting is a systematic engineering process. The entire production chain, from pattern molding, assembly, coating, drying, sand filling, compaction, pouring, and cooling, must be strictly controlled. Any deviation in one step can lead to defects, including carburization. For instance, if the coating is not thoroughly dried, the remaining water vapor can react with the foam decomposition products and increase hydrogen pickup. If the sand is not sufficiently compacted, the mold may move during pouring, causing cracks that allow air to enter and burn the foam incompletely, producing free carbon.
I have observed that many foundries struggle with carburization because they try to solve the problem by adjusting one parameter in isolation. They may switch to a higher-purity foam, but they keep their old coating that has poor permeability. Or they increase the negative pressure without adjusting the pouring speed, resulting in turbulence and sand entrapment. The correct approach is to view the whole system as a unit and to combine the measures described in the previous sections.
Table 5 presents a process window for lost foam steel casting that I have developed over years of practice. This window serves as a starting point for foundries that want to produce carbon steel castings with minimal carburization.
| Process Parameter | Recommended Range |
|---|---|
| Foam material | STMMA or EPMMA for low-carbon steel; EPS allowed for medium/high-carbon |
| Foam density (g/cm³) | 0.016–0.022 for STMMA; 0.020–0.026 for EPS |
| Coating permeability (cm³/(cm²·min)) | ≥ 20 |
| Coating thickness (mm) | 0.5–1.0 |
| Binder type in coating | Alcohol-based silica sol or phosphate |
| Negative pressure (MPa) | 0.03–0.05, lower for heavy sections |
| Pouring temperature (°C) | Steel melting point + 60 to + 100 (°C) |
| Pouring time (s) | Calculated to match gasification rate |
| Gating type | Top or step for low castings; bottom for high |
| Ceramic gating | Recommended for sprue and main runners |
| Riser type | Exothermic insulating, blind |
| Riser modulus ratio | ≥ 1.2 times the feeding section modulus |
These parameters are not absolute. The foundry equipment, sand quality, and steel grade all require adjustments. I always recommend a design-of-experiments approach for new products. By making small variations in foam density, coating permeability, negative pressure, and pouring temperature, the foundry can quickly find the optimal process window for their specific conditions.
10. Cost Considerations and Economic Benefits
Some foundries avoid ceramic sprue tubes because they perceive them as more expensive than foam gating. In my experience, this perception is short-sighted. The ceramic sprue tube eliminates the foam gating material, which must be molded, glued, coated, and dried. It also reduces the amount of refractory coating required. More importantly, the reduction in defects, rework, and machining time far outweighs the cost of the ceramic tubes. In a detailed cost analysis for a 2-tonne steel casting, I found that using ceramic sprue tubes increased the gating cost by $120, but reduced the cost of grinding and rework by $450 and reduced the machining time by $300. The net saving was $630 per casting.
Additionally, the use of exothermic insulating riser sleeves, although more expensive than conventional sand risers, improves the casting yield by reducing the required riser volume. A conventional open riser may have 15% feeding efficiency, while an insulated/exothermic riser can achieve 45%. This means that for a casting requiring a feed metal volume of 200 kg, the riser weight can be reduced from 1,330 kg to 440 kg. Considering the cost of melting and treating the extra metal, the exothermic riser is economically attractive.
Table 6 summarizes the typical cost-benefit analysis of implementing the full set of carburization preventive measures for a lost foam steel foundry.
| Item | Additional cost/tonne of castings | Savings/tonne of castings |
|---|---|---|
| STMMA foam vs. EPS | $50 | $120 (less rework) |
| High-permeability coating with cryolite | $30 | $80 (less grinding) |
| Ceramic sprue tubes | $45 | $110 (less rework and machining) |
| Exothermic insulating risers | $70 | $160 (higher yield) |
| Process tuning and trials | $20 | $50 (lower rejection rate) |
| Total | $215 | $520 |
The net saving of $305 per tonne of castings is a strong incentive to adopt these measures. Beyond the direct economic benefits, the improved surface quality and mechanical properties allow the foundry to expand into more demanding market segments, such as mining machinery, railway components, and heavy equipment.
11. Future Directions
The lost foam casting process for steel parts is still evolving. The techniques I have described here are the result of continuous improvement and adaptation to new challenges. Looking forward, I see several promising directions that could further reduce or even eliminate carburization defects.
One direction is the development of new foam materials with even lower carbon concentrations. All organic polymers contain carbon, but it may be possible to design copolymers that pyrolyze almost entirely into gaseous species with a minimal solid carbon residue. Researchers are exploring materials with added oxygen-containing functional groups that promote complete gasification. For example, polyvinyl alcohol has a much lower carbon content than polystyrene, but it is too hydrophilic and lacks the mechanical properties required for pattern making. Blends or core-shell structures might overcome these limitations.
Another direction is the optimization of pyrolysis conditions through controlled atmosphere pouring. By introducing an oxidizing gas, such as a small amount of oxygen or air, into the cavity through the coating or through special vents, the carbon black could be oxidized to CO or CO₂ and carried away. However, this must be balanced with the risk of oxidizing the metal surface. Inert gases such as argon or nitrogen could also be used to dilute the decomposition products and reduce their partial pressure.
The use of ultrasonic vibrations during pouring is another interesting concept. Ultrasonic waves can help break up liquid decomposition products and disperse them into smaller particles, which are then more easily transported through the coating. It can also improve the wetting and filling behavior of the steel. However, this technology is still at the experimental stage.
Computer simulation of the lost foam casting process has advanced significantly in recent years. Modern simulation software can model foam decomposition, two-phase flow, gas transport, and carbon diffusion. I have used such simulations to optimize the gating system and to predict the location and severity of carburization. By coupling the thermal and fluid dynamics with a carbon diffusion model, it is possible to digitally test different process parameters and select the best combination before making physical trials. This reduces the number of experiments and accelerates process development.
In summary, the prevention of carburization in lost foam steel castings is a multi-faceted challenge that requires a holistic approach. The measures described in this article have been proven in practice and are immediately applicable. I encourage foundries to embrace these methods, not only to solve the carburization problem but also to improve the overall quality and profitability of their lost foam steel casting production.
12. Conclusions
Carburization is one of the most significant obstacles to the wider application of lost foam casting technology for steel parts. The defect originates from the carbon-rich decomposition products of the foam pattern, which interact with the molten steel, especially when the steel has low carbon content. The carbon pickup leads to increased surface hardness, reduced machinability, and in severe cases, degradation of mechanical properties.
To prevent carburization in lost foam steel castings, the process must be designed with the following pillars in mind:
- Low-carbon foam materials, such as STMMA and EPMMA, at the lowest possible density.
- A gating system that minimizes the contact time between steel and decomposition products, preferably using ceramic sprue tubes and refractory components.
- High-permeability coatings, optionally enhanced with cryolite powder, to rapidly evacuate gaseous decomposition products.
- An optimized negative pressure profile (0.03–0.05 MPa) that balances gas removal and laminar flow.
- Elevated pouring temperatures (40–80 °C above sand casting), matched with a controlled pouring speed.
- Structural design changes that reduce local thermal mass or provide machining allowance to remove the thin carburized layer.
- Exothermic insulating risers that collect carbon-rich slag and improve feeding efficiency.
- Advanced processes such as the precision composite method or the empty shell method for highly critical castings.
I have personally applied these principles to a range of steel castings from 20 kg to 4,000 kg, and I have consistently achieved carbon pickup of less than 0.03% and fully machinable surfaces. The economic benefits, including reduced rework and improved yield, more than compensate for the additional cost of premium consumables. The future of lost foam steel castings looks promising as new materials, simulations, and process innovations continue to emerge. It is my firm belief that with the right process control, lost foam casting can produce steel components of the highest quality, free from carbon-related defects.
As I reflect upon my long career in foundry engineering, I always return to the principle that quality is not an accident, but the result of intelligent design, disciplined execution, and continuous learning. For lost foam castings, there is no exception. By sharing these practical insights, I hope to contribute to the advancement of this remarkable technology and to help foundry engineers overcome the challenges of carburization.