Lost foam casting (LFC) has gained significant attention in the foundry industry due to its near-net-shape capability, design flexibility, and environmental friendliness. However, the production of steel castings via the lost foam process still faces several challenges, among which the low feeding efficiency of risers is one of the most critical issues. Steel castings inherently exhibit large volumetric shrinkage during solidification, requiring a substantial amount of liquid metal to compensate for the contraction. In lost foam casting, the decomposition of the foam pattern absorbs a considerable amount of heat, and the negative pressure applied during pouring enhances the cooling effect at the metal–mold interface. Moreover, most risers used in lost foam steel casting are blind risers, which are buried in sand and tend to form a solidified skin on their upper surface during cooling. This skin prevents the atmospheric pressure from acting on the molten metal inside the riser, leaving only the metallostatic pressure of the liquid column to feed the casting. As a result, the feeding efficiency of blind risers is significantly reduced, leading to large shrinkage cavities and porosity in the final steel castings.
The present work aims to investigate and improve the riser feeding efficiency of lost foam steel castings through several approaches, including the use of insulating and exothermic riser sleeves, the application of atmospheric pressure via vent tubes, and the development of a novel exothermic cone insert placed at the top of blind risers. The study combines industrial-scale pouring experiments, numerical simulation using casting simulation software, differential scanning calorimetry (DSC) analysis, and metallographic examination to evaluate the effectiveness of the proposed methods.
1. Introduction and Motivation
The riser, also known as a feeder, is an essential element in casting design. Its primary function is to store liquid metal and feed the casting during solidification to compensate for volumetric shrinkage. For steel castings, the solidification shrinkage is large, typically in the range of 3–5% by volume. If the riser is not adequately designed or if its feeding efficiency is low, the casting will develop shrinkage defects. The feeding efficiency of a riser is defined as the ratio of the amount of metal delivered to the casting to the total volume of the riser. In conventional sand casting, well-designed risers can achieve feeding efficiencies of 10–15%. However, in lost foam steel casting, the feeding efficiency can drop to less than 6% because of the early skin formation at the riser top and the heat absorption by the foam pattern.
During my investigation of dozens of lost foam steel casting foundries, I observed that most designers rely on empirical rules and oversized risers to compensate for the poor feeding behavior. The typical yield (casting weight divided by total poured weight) is only 50–60%, which inevitably increases production cost. Therefore, improving the riser feeding efficiency is of paramount importance for the economic viability of lost foam steel castings.
2. Riser Types and Design Principles for Lost Foam Steel Castings
2.1 Classification of Risers
Risers can be classified according to their position, top condition, and pressurization method. The main types are listed in the following table:
| Classification criterion | Riser type | Description |
|---|---|---|
| Position | Top riser | Placed directly above the casting hot spot |
| Position | Side riser | Placed adjacent to the casting, connected through a riser neck |
| Top condition | Open riser | Exposed to atmosphere, easy to add exothermic powder |
| Top condition | Blind riser | Fully enclosed in the mold, better heat retention but prone to skin formation |
| Pressurization | Atmospheric riser | Vented to atmosphere through a core or tube to enhance feeding pressure |
| Pressurization | Exothermic riser | Contains exothermic material that reacts with the metal to release heat |
| Pressurization | Insulated riser | Wrapped with insulating sleeve to reduce heat loss |
In lost foam casting, blind risers are widely used because they occupy less space and can be placed at any position. However, as mentioned, they suffer from early crust formation. The shape of the riser also influences its feeding behavior. For the same volume, a sphere has the smallest surface area and therefore the longest solidification time. The following table shows solidification times for identical weights of various riser shapes in steel casting:
| Riser shape | Dimensions (mm) | Surface area (m²) | Solidification time (min) |
|---|---|---|---|
| Sphere | Diameter = 152 | 2.54 | 7.2 |
| Cylinder | Diameter = 108, Height = 203 | 3.05 | 4.7 |
| Rectangular block | 79 × 92 × 219 | 3.43 | 3.6 |
Thus, cylindrical risers with a slight taper are commonly used to combine ease of molding with acceptable feeding efficiency.
2.2 Riser Size Calculation Methods
Several methods exist for calculating riser dimensions. The most common ones are:
- Proportional method: The riser diameter is taken as a multiple of the hot spot circle diameter. For steel castings, the ratio is often expressed as \( D = C d \), where \( D \) is riser diameter, \( d \) is the inscribed circle diameter of the hot spot, and \( C \) is an empirical coefficient (usually 1.2–1.8).
- Formula method: Based on the solidification time ratio and volumetric shrinkage, a general riser equation is derived:
\[
\frac{V_r}{V_c} = \frac{A_r \cdot f_s}{A_c \cdot f_c – A_r \cdot f_s}
\]
where \( V_r, V_c \) are volumes, \( A_r, A_c \) are surface areas, and \( f_s, f_c \) are solidification factors of riser and casting, respectively.
- Modulus method: The modulus \( M \) is defined as the ratio of volume to cooling surface area:
\[
M = \frac{V}{S}
\]
For a riser to feed effectively, the riser modulus must be larger than the casting modulus. For steel castings, the following empirical relations are often used:
\[
M_{\text{riser}} \geq 1.2 M_{\text{casting}} \quad (\text{top open riser})
\]
\[
M_{\text{riser neck}} \geq 1.1 M_{\text{casting}}
\]
In addition to the modulus condition, the riser must contain enough liquid metal to compensate for the solidification shrinkage. The required riser volume can be checked by:
\[
V_r \geq \varepsilon \cdot V_c \cdot \frac{1}{\eta}
\]
where \( \varepsilon \) is the volumetric shrinkage, \( V_c \) is the casting volume, and \( \eta \) is the riser feeding efficiency. Typical efficiency values for various riser types are listed below:
| Riser type | Feeding efficiency η (%) |
|---|---|
| Cylindrical or kidney-shaped | 12–15 |
| Spherical | 15–20 |
| With top pouring (direct pour) | 15–20 |
| Exothermic / insulated | 25–30 |
| Atmospheric pressure | 15–20 |
| Compressed air | 35–40 |
3. Methods to Improve Riser Feeding Efficiency for Lost Foam Steel Castings
Based on the fundamentals of solidification, the feeding efficiency can be enhanced by either (a) increasing the feeding pressure acting on the molten metal, or (b) prolonging the liquid state of the metal inside the riser. The following methods were explored in this study.
3.1 Insulating and Exothermic Riser Sleeves
Insulating sleeves are made of materials with low thermal conductivity, such as vermiculite, perlite, and ceramic fiber. Exothermic sleeves contain aluminum powder and oxidizers (e.g., iron oxide) that react when in contact with molten steel, releasing heat. The combination of insulation and exothermic action effectively increases the modulus of the riser. The modulus enhancement factor \( E \) is defined as:
\[
M_{\text{exo}} = E \cdot M_{\text{normal}}
\]
where \( E \) typically ranges from 1.3 to 1.6 for commercial insulating sleeves, and may reach 3–4 for exothermic sleeves. The relationship between sleeve thickness \( \delta \) and \( E \) is not linear; beyond a certain thickness, the benefit diminishes.
In this study, I applied a commercial exothermic riser sleeve to a small crankshaft steel casting (45# steel, 21 kg) produced by lost foam. One casting used the traditional foam riser, while another identical casting used the exothermic sleeve. After pouring and cooling, both risers were cut open for comparison. The ordinary riser weighed 11.2 kg, and the casting yield was only 52%. The exothermic-sleeved riser weighed only 4.8 kg, raising the yield to 73%. The shrinkage cavity in the insulated riser was concentrated in the upper part, indicating that the riser could be further downsized.
3.2 Application to a Large Hollow Steel Ball
For larger castings, insulating sleeves also demonstrate significant benefits. I collaborated with a foundry to produce a wear-resistant hollow steel ball used in an E-type coal mill. The ball had an outer diameter of 1150 mm, inner diameter of 850 mm, and a mass of 3600 kg. The material was ZG75Cr2MnNiMo. The original process used two large ordinary risers with dimensions \( D=482 \) mm and \( H=725 \) mm, resulting in a yield of 63%. By applying insulating riser sleeves, the riser size was reduced to \( D=370 \) mm and \( H=562 \) mm. Two pouring system designs were evaluated: side gating (Scheme 1) and top gating through the riser (Scheme 2). Numerical simulation was performed using the HuaZhu CAE software to predict shrinkage porosity and compare processes.
The simulation parameters for ZG75Cr2MnNiMo are listed below:
| Property | Value |
|---|---|
| Initial temperature (°C) | 1560 |
| Density (g/cm³) | 7.08 |
| Specific heat (cal/g·°C) | 0.3168 |
| Thermal conductivity (cal/cm·s·°C) | 0.082 |
| Viscosity (cm²/s) | 0.04 |
| Radiation coefficient | 0.375 |
| Latent heat (cal/g) | 53.9 |
| Liquidus temperature (°C) | 1470 |
| Solidus temperature (°C) | 1368 |
| Critical solid fraction | 0.68 |
| Solidification coefficient | 1.3 |
| Liquid volumetric shrinkage (1/°C) | 2.75×10⁻⁵ |
The simulation results showed that the top-gated scheme (Scheme 2) produced fewer shrinkage defects than the side-gated scheme, because the riser received hot metal directly during pouring, maintaining a higher temperature for a longer period. The yields were 73.6% for Scheme 1 and 76.2% for Scheme 2, compared to 63% for the original process. Therefore, the combined use of insulating sleeves and direct pouring through the riser can significantly improve riser feeding efficiency and reduce steel consumption.
3.3 Atmospheric Pressure Riser
Another effective method is to create an atmospheric pressure riser by inserting a refractory core or tube through the top of the blind riser, allowing the atmosphere to communicate with the interior of the riser after the top surface has solidified. The atmospheric pressure acting on the molten metal increases the feeding potential. The theoretical maximum height that can be fed by atmospheric pressure for steel castings is calculated as:
\[
H = \frac{760 \times \rho_{\text{Hg}}}{\rho_{\text{steel}}}
\]
where \( \rho_{\text{Hg}} = 13.6 \) g/cm³ and \( \rho_{\text{steel}} = 7.8 \) g/cm³, giving \( H \approx 1325 \) mm. However, in practice, due to inter-dendritic flow resistance, the effective feeding distance is much shorter.
In lost foam casting, I implemented a simple version by gluing a ceramic tube onto the top of the foam riser before coating and molding. The tube extended above the sand and was covered with foam during pouring. After pouring, the foam plug was burned off, connecting the riser interior to the atmosphere. This modification effectively eliminated the skin formation problem and significantly improved the feeding action.
4. Design and Testing of an Exothermic Cone Insert
Inspired by the Williams core concept and the principles of exothermic risers, I designed a conical insert made of an exothermic mixture. The cone is placed at the top of a blind riser, with its apex pointing downward. When the molten steel contacts the cone, an aluminothermic reaction occurs, releasing a large amount of heat. This heat delays the solidification of the metal at the riser top and keeps a liquid channel open, thereby improving the feeding efficiency. Because the cone is much smaller than a full exothermic sleeve, its material cost is lower and it is easier to handle.
4.1 Cone Geometry and Materials
Three sizes of exothermic cones were designed:
| Size | D₁ (mm) | D₂ (mm) | H₁ (mm) | H₂ (mm) | Price (RMB) |
|---|---|---|---|---|---|
| Small | 50 | 30 | 20 | 8 | 0.9 |
| Medium | 60 | 40 | 40 | 12 | 1.4 |
| Large | 100 | 60 | 65 | 15 | 5.6 |
The exothermic mixture consists of aluminum powder, iron oxide (Fe₂O₃ or Fe₃O₄), coal powder, magnesium powder as an igniter, water glass as a binder, and water. The main exothermic reactions are:
\[
8Al + 3Fe_3O_4 \rightarrow 4Al_2O_3 + 9Fe + \Delta H
\]
\[
2Al + Fe_2O_3 \rightarrow Al_2O_3 + 2Fe + \Delta H
\]
The reaction starts when the local temperature exceeds approximately 1000°C. The heat released further heats the surrounding metal, preventing premature solidification.
4.2 Differential Scanning Calorimetry (DSC) Analysis
To verify that the exothermic cone can release heat in the low-oxygen environment typical of lost foam casting (where negative pressure is applied), I conducted DSC experiments on the cone material both in air and under an argon atmosphere. Figure 1 shows the DSC curves. Although the curves differ in magnitude, endothermic/exothermic peaks appear at approximately 130°C, 320°C, and 420°C in both cases. The major exothermic peak corresponding to the aluminothermic reaction occurs above 1000°C, which is not visible in the DSC scan limited to 600°C, but the preliminary peaks indicate that the mixture is reactive even without external oxygen. Therefore, the cone can function inside a blind riser during lost foam pouring.
4.3 Pouring Experiments with the Small Cone
A series of comparative experiments was carried out on a small alloy hammer head (ZG30CrMnSiBRE, weight 7 kg). Three riser configurations were tested under identical pouring conditions:
- Group A: ordinary blind riser (no insert)
- Group B: blind riser with the small exothermic cone
- Group C: blind riser with the cone plus a ceramic vent tube to atmosphere
The procedure involved cutting a conical recess in the top of the foam riser, inserting the cone, and sealing it with glue and coating. For Group C, a ceramic tube was attached to the cone top and passed through the sand and the plastic film at the top of the mold. The tube was plugged with foam during pouring, then the foam was burned off immediately after pouring to allow atmospheric pressure to act.
After solidification and cooling, the risers were sectioned and the shrinkage cavities were measured using a liquid water injection method. The total shrinkage volume for each riser is given below:
| Riser type | Shrinkage volume (cm³) |
|---|---|
| Ordinary blind riser | 10.7 |
| With exothermic cone | 12.3 |
| With cone + vent tube | 13.2 |
The increasing shrinkage volume indicates that more metal was transferred from the riser to the casting, i.e., the feeding efficiency was improved. The casting with the ordinary riser showed visible shrinkage at the hot spot, while those with the cone or cone+tube were sound. Moreover, the density of the riser sections was measured to verify that the water-injection method captured most of the internal porosity. The density differences were small, confirming the reliability of the measurement.
4.4 Application to a Balance Shaft Housing
To confirm the effectiveness of the small cone on a larger and more complex casting, I also tested it on a balance shaft housing (ZG310-570, mass 50 kg) produced by a cooperating foundry. The casting was poured in a ganged layout with a bottom gating system, and the risers were located far from the sprue to simulate cold riser conditions. The same three configurations were used. After sectioning, the ordinary riser showed dispersed shrinkage throughout its lower half, while the riser with the cone had a concentrated deep shrinkage cavity in the upper part, with a clean and dense lower region. The measured volumes were:
| Riser type | Shrinkage volume (cm³) |
|---|---|
| Ordinary blind riser | 52 |
| With exothermic cone | 144 |
| With cone + vent tube | 148 |
This experiment clearly demonstrated that the cone significantly improved the feeding efficiency, especially for risers that are not directly connected to the ingate. The concentrated shrinkage in the upper riser also implies that the riser height could be reduced, further increasing the casting yield.
4.5 Application of the Large Cone to a Medium-Sized Hammer Head
A medium-sized hammer head (40Cr steel, 186 kg) was used to evaluate the large exothermic cone. The same three configurations were tested. After casting, the top surfaces of the risers were observed. The ordinary riser exhibited a honeycomb-like top surface, indicating that it had not completely crusted in this particular case. The risers with the cone showed larger open cavities. The measured shrinkage volumes were:
| Riser type | Shrinkage volume (cm³) |
|---|---|
| Ordinary blind riser | 500 |
| With exothermic cone | 556 |
| With cone + vent tube | 562 |
Although the differences were less pronounced than in the previous experiments, the cone still provided a measurable improvement. The results suggest that the benefits of the cone are more significant when the riser would otherwise be a cold riser or when the riser is located in a lower part of the mold where the metallostatic head is low.
4.6 Recommended Application Ranges for Exothermic Cones
Based on the experiments, I have developed preliminary guidelines for the selection of the appropriate cone size:
| Cone size | Suitable riser diameter (mm) | Suitable riser height (mm) | Riser metal weight (kg) | Max casting weight (kg) |
|---|---|---|---|---|
| Small | 80–130 | 100–200 | 5–20 | 100 |
| Medium | 120–210 | 180–350 | 15–30 | 300 |
| Large | 200–450 | 320–500 | 30–180 | 800 |
These ranges are based on a limited number of trials and may need further refinement as more data become available.
4.7 Metallographic Examination
To investigate whether the exothermic cone material contaminates the steel castings, I sampled specimens from the bottom of each riser (near the casting interface) for metallographic analysis. The material was ZG30CrMnSiBRE in the as-cast condition, etched with 8% nitric acid in alcohol. The microstructures observed under an optical microscope are summarized below:
| Sample | Microstructure |
|---|---|
| Ordinary riser bottom | Pearlite matrix with white acicular/blocky ferrite (Widmanstätten structure) plus minor oxide inclusions |
| Riser with cone | Similar pearlite + ferrite structure, with comparable inclusion content |
| Riser with cone + vent tube | Similar microstructure, no significant differences |
The presence of oxide inclusions was similar in all samples, indicating that the exothermic cone does not introduce additional harmful inclusions to the casting. Because the cone material has a very low density (approximately 0.75 g/cm³), any fragmented cone pieces tend to float to the riser top and remain in the shrinkage cavity rather than entering the casting.
5. Discussion on Numerical Simulation of Lost Foam Steel Castings
Computer simulation is a powerful tool for optimizing riser design and predicting shrinkage defects. In this work, the HuaZhu CAE software was used to simulate the filling and solidification of the large hollow steel ball. The simulation allowed a direct comparison of the side-gated and top-gated designs, as well as an evaluation of the insulating sleeve effects. However, it is important to note that lost foam casting involves complex physical and chemical phenomena, including foam decomposition, gas evolution, coating permeability, and the influence of negative pressure. Most commercial simulation packages model the filling as a free-surface flow with heat transfer but often simplify or ignore the detailed foam–metal interaction. Consequently, some discrepancies may exist between simulation predictions and actual castings. Nevertheless, simulation remains a valuable tool for relative comparisons and for identifying trends. The following table summarizes the key simulation results for the steel ball:
| Process | Riser weight (kg) | Gating weight (kg) | Yield (%) | Shrinkage porosity level |
|---|---|---|---|---|
| Original (ordinary risers) | 1933 | 182 | 63.0 | Higher |
| Insulating sleeve, side gate | 946 | 341 | 73.6 | Medium |
| Insulating sleeve, top gate through riser | 946 | 176 | 76.2 | Lower |
The simulation results clearly indicate that top gating through the riser is beneficial for feeding, as it ensures that the riser remains the hottest part of the system. This finding aligns with the practical recommendation to use direct pouring (i.e., the riser serves as the ingate) whenever possible.
6. Conclusions and Outlook
The main conclusions derived from this work can be summarized as follows:
- Insulating and exothermic riser sleeves can significantly enhance the feeding efficiency of lost foam steel castings by increasing the effective modulus of the riser. In the case of the crankshaft, the use of an exothermic sleeve improved the casting yield from 52% to 73%.
- Computer simulation is a useful tool for optimizing the riser and gating system design for lost foam steel castings. The simulation of a large hollow steel ball confirmed that a top-gated system with insulating sleeves provides better feeding and a higher yield than a side-gated system with larger ordinary risers.
- The newly developed exothermic cone, inserted at the top of a blind riser, is a low-cost and easy-to-use solution to prevent early skin formation and improve feeding efficiency. Comparative pouring experiments on three different steel castings demonstrated that the cone increases the shrinkage cavity volume in the riser, which is a direct indication of improved metal feeding to the casting.
- Adding a vent tube connecting the riser top to the atmosphere further enhances the feeding effect by allowing atmospheric pressure to act on the molten metal. The best results were obtained when the vent tube was combined with the exothermic cone.
- The exothermic cone does not have a detrimental effect on the casting material. Metallographic examination confirmed that the microstructure and inclusion content of riser bottoms from cone-equipped and ordinary risers are comparable.
- For risers that are already directly connected to the ingate (hot risers), the additional benefit of the cone may be limited because the hot metal flowing through the riser naturally maintains a high temperature. However, for cold risers or those located far from the sprue, the cone can provide substantial improvements.
In the future, several aspects warrant further investigation:
- The influence of the exothermic cone on carbon pickup in low-carbon steel castings should be studied, since the cone contains organic binders and carbonaceous materials that might interact with the molten steel.
- Optimization of the cone composition to minimize gas generation while maximizing heat release would be beneficial, especially for high-quality steel castings that are sensitive to gas porosity.
- More extensive industrial trials are needed to refine the recommended application ranges for the different cone sizes and to establish a comprehensive database for different casting geometries and steel grades.
- The behavior of the cone under varying negative pressure levels and coating systems should be characterized to provide more precise design guidelines.
In conclusion, improving riser feeding efficiency in lost foam steel castings is a multi-faceted problem that can be addressed through a combination of proper riser design, the use of insulating/exothermic materials, atmospheric pressure assistance, and optimized gating systems. The exothermic cone proposed in this work offers a practical and economical solution that can help foundries reduce metal waste, lower energy consumption, and enhance the soundness of steel castings.
As the foundry industry moves toward more sustainable and cost-effective manufacturing, continued research on riser optimization and innovative feeding aids will play a crucial role in expanding the application of lost foam technology for steel castings.