In modern tractor manufacturing, the divided axle housing is one of the most critical load-bearing components. It must withstand high impact loads under full-load conditions across diverse terrains such as plateaus, saline-alkali lands, and deserts. Consequently, the casting quality of the axle housing is of paramount importance. My work focuses on a specific type of divided axle housing produced in a sand casting foundry. In this foundry, the annual production volume is approximately 10,000 units, but the occurrence rate of shrinkage porosity and shrinkage cavity defects was as high as 70%, leading to oil leakage during service and annual economic losses exceeding one million RMB. To fundamentally solve this problem, I employed the ProCAST finite element simulation software to analyze the mold filling and solidification processes, designed three different gating systems, and optimized the casting process. This research demonstrates how modern simulation tools can profoundly improve the quality and yield of sand casting foundry operations.
1. Background and Research Significance
The divided axle housing is mainly manufactured by sand casting foundry processes using sodium silicate-bonded sand molds. The original casting process adopted a bottom gating system, which caused poor filling capacity and serious shrinkage defects located near the inner semi-axle oil seal seat. Through systematic defect analysis, I found that the solidification process inside the mold created isolated liquid pools in this critical region, preventing simultaneous solidification and resulting in internal cavities. Since such defects are invisible to the naked eye and difficult to detect by conventional methods, they often lead to serious quality accidents during service.
The main objective of this research is to design a robust casting system for the divided axle housing, to verify the design through numerical simulation, and to optimize the process by adding risers and chills. The key phrase “sand casting foundry” represents both the physical production site and the methodological framework within which this study is conducted. By integrating simulation with experimental validation, I aim to establish a reliable reference for similar sand casting foundry problems.
In the past few decades, casting simulation technology has evolved from simple temperature field calculations to comprehensive coupled analysis of flow, heat transfer, and stress. Tools like ProCAST allow engineers to visualize the filling sequence, temperature distribution, and defect formation in a virtual environment. In our sand casting foundry, this capability has become indispensable for reducing trial-and-error time and improving product quality.
The above photograph illustrates a typical sand casting foundry environment, where molds are prepared, molten metal is poured, and castings are solidified. This practical context underpins the entire research presented in this paper.
2. Structural and Process Analysis of the Axle Housing
The divided axle housing under investigation has a complex geometry with significant wall thickness variations. The main external dimensions are 452 mm × 340 mm × 396 mm, with an average wall thickness of 25 mm. The maximum wall thickness reaches 29 mm, while the minimum is only 15 mm. The component is made of ZG40Mn low-alloy cast steel, which offers excellent strength and toughness after proper heat treatment. The chemical composition and mechanical properties are listed in Tables 2-1 and 2-2.
| Element | C | Mn | Si | S | P |
|---|---|---|---|---|---|
| Content | 0.35–0.45 | 1.0–1.5 | 0.35–0.45 | ≤0.03 | ≤0.03 |
| σs (MPa) | σb (MPa) | δ5 (%) | ψ (%) | Hardness (HB) |
|---|---|---|---|---|
| ≥295 | ≥640 | ≥12 | ≥30 | 163–200 |
The raw materials used in the sand casting foundry include scrap steel, ferrosilicon, high-carbon ferromanganese, and silicon-aluminum-iron alloy. Strict quality control of these materials is essential to avoid unwanted impurities that can promote shrinkage defects. The molding sand is sodium silicate-bonded quartz sand with suitable particle shape and size distribution. For this axle housing, I selected CO₂-hardened sodium silicate sand, which provides adequate strength and excellent collapsibility.
In analyzing the structural processability, I determined that the casting position should be horizontal, with the parting surface located at the middle plane of the housing. This arrangement facilitates mold production, core setting, and ensures that the important machined surfaces are properly filled. The pouring temperature was set to 1580 °C based on the need to balance fluidity and solidification behavior. The casting tolerance grade was chosen as CT12 according to GB6414-1999, and the linear shrinkage coefficient was determined to be 2%.
The core design is crucial because the housing contains a continuous internal cavity. I designed a single sand core with both horizontal and vertical core prints. The horizontal core prints have a length of 50 mm on each side with a taper of 3°, while the vertical upper core print is 50 mm long with a 4° taper, and the lower core print is 30 mm long with a 1° taper. The pattern draft angle was set to 3° according to the sand casting foundry empirical guidelines.
3. Design of the Gating System
The gating system must ensure that molten steel fills the mold quickly and smoothly, avoids oxidation and slag entrapment, and promotes directional solidification. For steel castings, the gating system should be simple with large cross-sections. The pouring temperature is higher than for iron castings, so the refractory quality of the mold is critical.
I first determined the ingate position. Among various options, the middle gating system (mid-riser) was chosen because it combines the advantages of top and bottom gating. The minimum residual pressure head was calculated using Equation (3-1):
$$ h_M = L \tan \alpha \tag{3-1} $$
where \( h_M \) is the minimum remaining pressure head, \( L \) is the flow distance from the sprue centerline to the highest point of the casting, and \( \alpha \) is the pressure angle. For this casting, \( L = 320 \) mm, \( \alpha = 10^\circ \), so \( h_M = 320 \times \tan 10^\circ \approx 56 \) mm. The sprue height must be at least 56 mm + 190 mm = 246 mm.
Using the pouring rate method, the choke area was calculated from Equation (3-2):
$$ A_{\text{阻}} = \frac{G_L}{t \cdot k \cdot S’} \tag{3-2} $$
where \( G_L \) is the weight of liquid metal flowing through the choke, \( t \) is the pouring time, \( k \) is the casting speed factor, and \( S’ \) is the fluidity coefficient (0.8 for high-manganese steel). The weight of the casting was calculated from its volume (approx. 15,000 cm³) and steel density (7.8 g/cm³), giving about 117 kg. The relative density \( \rho = G_L / V_C = 108 / (45 \times 34 \times 40) \approx 0.002 \) kg/cm³. From standard tables, \( C = 0.8 \) and \( k = 0.6 \) kg/(cm²·s). The pouring time was found as \( t = C \sqrt{G_L} = 0.8 \times \sqrt{108} \approx 8.3 \) s, but I used 10 s for practical reasons. Substituting into Eq. (3-2): \( A_{\text{阻}} = 108 / (10 \times 0.6 \times 0.8) = 22.5 \) cm².
Based on the choke area, I designed three different gating system ratios:
- Scheme 1 – Closed system: \(\sum A_{\text{in}} : \sum A_{\text{run}} : \sum A_{\text{sprue}} = 1 : 1.3 : 1.5\)
- Scheme 2 – Open system: \(\sum A_{\text{in}} : \sum A_{\text{run}} : \sum A_{\text{sprue}} = 1.2 : 1.1 : 1\)
- Scheme 3 – Semi-closed system: \(\sum A_{\text{in}} : \sum A_{\text{run}} : \sum A_{\text{sprue}} = 1 : 0.8 : 1.2\)
For each scheme, I calculated the dimensions of the sprue, runner, and ingate. Table 3-4 summarizes the three designs.
| Scheme | Ingate area (cm²) | Runner area (cm²) | Sprue area (cm²) | Ratio |
|---|---|---|---|---|
| Scheme 1 closed | 22.5 | 29.25 | 33.75 | 1:1.3:1.5 |
| Scheme 2 open | 22.5 | 21 | 18.8 | 1.2:1.1:1 |
| Scheme 3 semi-closed | 22.5 | 18 | 27 | 1:0.8:1.2 |
The ingate was designed as a cylindrical port to simplify cleaning and improve feeding efficiency. The runner had a trapezoidal cross-section, and the sprue was cylindrical. Additionally, a ceramic foam filter was placed at the bottom of the sprue to trap inclusions and stabilize the flow. These designs are representative of best practices in the sand casting foundry.
4. Numerical Simulation of Filling and Solidification
I used ProCAST finite element software to simulate the mold filling and solidification processes of the axle housing under the three gating schemes. The three-dimensional CAD model was created in Creo and imported into ProCAST, where surface and volume meshes were generated. Local refinement was applied to critical thin sections to ensure at least three elements across the wall thickness. The mesh element size was 6 mm for the casting and 15 mm for the gating system. A virtual mold was used for heat transfer calculations.
The thermal physical properties of ZG40Mn were taken from the ProCAST database. The liquidus temperature is approximately 1510 °C and the solidus around 1440 °C. The pouring temperature was set to 1580 °C, the initial mold temperature to 30 °C, and the interface heat transfer coefficient between casting and mold was set to 500 W/(m²·K). The gravity acceleration was 9.8 m/s². The pouring time was controlled at 10 seconds.
4.1 Filling Process Analysis
By examining the velocity fields during filling, I observed significant differences among the three schemes. For the closed system (Scheme 1), the molten metal flowed steadily through the gating system, but there was some splashing in the early stage due to the high velocity. The metal remained in a saturated state, providing good slag retention. For the open system (Scheme 2), the flow was unstable, with many isolated droplets and turbulence. The solidification fraction reached 18.8% at the end of filling, indicating premature solidification and a high risk of cold shuts. In contrast, the semi-closed system (Scheme 3) exhibited the smoothest flow with no dropout, and the mold was filled uniformly. Table 4-3 compares the solidification fractions at different filling stages.
| Scheme | Fill 25% | Fill 30% | Fill 64% | Fill 96% |
|---|---|---|---|---|
| Scheme 1 (closed) | 0 | 0 | 2.5 | 10.4 |
| Scheme 2 (open) | 0 | 0.1 | 10.5 | 18.8 |
| Scheme 3 (semi-closed) | 0 | 0 | 4 | 12.7 |
4.2 Temperature Field and Shrinkage Defect Prediction
The temperature field simulations allowed me to visualize the solidification sequence. In all initial schemes, isolated liquid pools appeared in the thick sections during the final stage of solidification, creating a high risk of shrinkage cavities. The open system (Scheme 2) produced the largest isolated regions and the most extensive defects. Based on these results, I eliminated Scheme 2 from consideration.
For the remaining two schemes (closed and semi-closed), I introduced risers and chills to improve feeding. The riser diameter was calculated using Equation (4-3):
$$ D_r = (1.2 \sim 2) \sigma_c \tag{4-3} $$
$$ H_r = (1.0 \sim 1.5) D_r \tag{4-4} $$
where \( \sigma_c \) is the local hot spot thickness (22 mm). For this case I used \( D_r = 80 \) mm and \( H_r = 150 \) mm. The modulus of the riser was \( M_{\text{riser}} = V/A \approx 18.75 \) mm, which is larger than the casting modulus of 9.7 mm, satisfying the requirement. Chills of dimensions 30 mm × 50 mm × 60 mm were placed in the internal cavity where shrinkage defects tended to form, because risers could not be positioned there. The chills increase the local cooling rate and promote directional solidification.
After adding risers and chills, I re-simulated both schemes. The results showed that the closed system still exhibited small shrinkage cavities near the flange end. However, the semi-closed system achieved complete elimination of internal defects; all porosity was transferred into the risers, and the casting became sound. Table 4-4 summarizes the simulation outcomes.
| Scheme | Flow stability | Final solidification fraction (%) | Defect prediction | After optim. | Conclusion |
|---|---|---|---|---|---|
| Scheme 1 (closed) | Good | 86 | Local shrinkage | Not fully eliminated | Rejected |
| Scheme 2 (open) | Poor | 92 | Large shrinkage | N/A | Rejected |
| Scheme 3 (semi-closed) | Excellent | 82 | Local shrinkage | Eliminated | Best scheme |
This demonstrates that the semi-closed gating system, combined with properly placed risers and chills, provides the most favorable solidification condition. The method described here is generally applicable to other similar steel castings produced in a sand casting foundry.
5. Experimental Validation
To verify the optimized semi-closed casting system, I conducted a series of production trials in the sand casting foundry. The pouring temperature was 1580 °C, the pouring time was 10 seconds, and the final gating system had a sprue area of 27 cm², a runner area of 18 cm², and an ingate area of 22.5 cm². The risers were arranged both as blind risers near the ingate and as open risers on the top surface. Chills were placed in the internal cavity and at thick-wall locations.
5.1 Sectioning Inspection
The first two castings from the initial batch were cut into sections at intervals of 100 mm. The cut surfaces were examined using a 10× magnifier. No visible cracks or any cavities larger than 3 mm were observed at the critical locations. Over five production batches, a total of 630 castings were produced, and 31 castings were sectioned for internal quality inspection. All samples met the requirement, showing no shrinkage-related defects.
5.2 Air Tightness Test
The air tightness test is essential for verifying that the casting is free from micro-porosity that could cause oil leakage. The machined axle housings were sealed with dedicated fixtures and immersed in water at a depth of 10–20 cm. Compressed air was gradually applied until the pressure reached 0.5 MPa. The dwell time was at least 10 minutes per part. Any emerging bubbles indicated a leak. Two batches, consisting of 8 and 12 parts respectively, were tested. The results are recorded in Table 5-1.
| Sample No. | Pressure (MPa) | Sealing condition | Time (min) | Leakage | Qualification |
|---|---|---|---|---|---|
| A1–A8 | 0.5 | Intact | 10 | No | Qualified |
| B1–B12 | 0.5 | Intact | 10 | No | Qualified |
All twenty tested housings passed the air tightness test. By the end of 2019, more than 6,000 units had been produced using the optimized process, and the nonconforming rate was reduced to below 3%, compared with the original 70% defect rate. This dramatic improvement confirms the effectiveness of the semi-closed gating system and the overall design methodology in the sand casting foundry environment.
6. Conclusions and Future Work
In this research, I performed a comprehensive study on the sand casting foundry process of a tractor divided axle housing. The main conclusions are as follows:
- The semi-closed gating system ( \(\sum A_{\text{in}} : \sum A_{\text{run}} : \sum A_{\text{sprue}} = 1 : 0.8 : 1.2\) ) was found to be the optimal choice for this casting. Both flow field and temperature field simulations indicated smooth filling, no entrapment, and favourable solidification conditions.
- Risers and chills are essential to eliminate shrinkage defects. The blind and open risers provided adequate feeding, while the chills promoted directional solidification. After optimization, all internal porosity was removed.
- Experimental production confirmed the simulation results. The reject rate was reduced from 70% to less than 3%, and all tested components passed the stringent air tightness test.
Future work should include coupling stress field simulation to predict hot tearing and residual deformation. Additionally, more accurate thermo-physical property data obtained from actual production experiments would enhance the reliability of the simulations. The methodology developed here can be readily extended to other large steel castings in the sand casting foundry industry.