Driven by the high-quality development of China’s new urbanization, the continuous expansion of urban agglomerations necessitates advanced transportation infrastructure. Urban rail transit, a key component, presents a sustained demand for critical components such as traction system motor housings. These housings are typically manufactured as castings, with nodular cast iron (ductile iron), specifically grade QT500-7, being a preferred material due to its excellent combination of strength, ductility, and castability. This article details the comprehensive process design, simulation-assisted development, and iterative refinement undertaken to successfully produce a qualified thin-walled nodular cast iron motor housing casting.
The subject component is a barrel-shaped structure with approximate overall dimensions of 760 mm in length, 500 mm in width, and 580 mm in height, weighing about 160 kg. Its defining characteristic is a predominant wall thickness of only 10 mm, juxtaposed with significantly thicker sections such as mounting brackets, reaching up to 90 mm. This variation in section size inherently creates isolated thermal centers, or hot spots, which are primary candidates for shrinkage porosity defects in nodular cast iron. The casting must meet stringent quality standards, requiring X-ray inspection with acceptance levels of Class 2 for critical areas and Class 3 for non-critical areas. The principal challenge lies in designing a casting process that ensures soundness in these isolated thick sections while filling the extensive thin walls completely.
The initial phase of process design involved a thorough analysis of the casting geometry to determine the optimal pouring position and parting line. Two primary orientations were considered: vertical (with the barrel axis perpendicular to the parting plane) and horizontal (with the barrel axis parallel to the parting plane). A horizontal parting, split along the barrel’s axis, was selected. This scheme, while presenting challenges for the internal ribs, offered significant advantages: it simplified molding by reducing the number of required cores compared to a vertical orientation, and it positioned major hot spots like the upper and lower mounting brackets favorably near the cope surface, facilitating the application of effective feeding mechanisms.
The gating system was designed as a horizontally parted, pressurized system with a single sprue, a runner bar along the parting plane adjacent to the barrel wall, and two ingates. Key parameters were calculated using established empirical formulas for nodular cast iron castings. The total pouring time \( t \) (in seconds) was estimated based on the total mass of metal \( G \) (in kg) in the mold:
$$ t = K \sqrt{G} $$
With \( G \approx 180 \, \text{kg} \) and the empirical coefficient \( K = 1.85 \), the calculated pouring time was \( t \approx 25 \, \text{s} \). The minimum choke cross-sectional area \( S_{\text{choke}} \) (in cm²) was determined to control the flow rate:
$$ S_{\text{choke}} = \frac{G}{0.31 \mu t \sqrt{H_p}} $$
where \( \mu \) is the total flow resistance coefficient (taken as 0.48) and \( H_p \) is the effective metallostatic pressure head (calculated as 35 cm). This yielded \( S_{\text{choke}} \approx 8.5 \, \text{cm}^2 \), which was subsequently increased to \( 12 \, \text{cm}^2 \) based on practical experience for thin-walled castings to ensure rapid filling. The cross-sectional areas of the gating system elements were set in a semi-pressurized ratio: \( \Sigma S_{\text{sprue}} : \Sigma S_{\text{runner}} : \Sigma S_{\text{ingate}} = 1.2 : 1.5 : 1 \). Consequently, a sprue diameter of 40 mm (\( S \approx 12.6 \, \text{cm}^2 \)), a total runner area of \( 18 \, \text{cm}^2 \), and two ingates each of \( 6 \, \text{cm}^2 \) were specified.
| Process Parameter | Symbol | Value | Unit |
|---|---|---|---|
| Total Casting Mass | \( G \) | ~180 | kg |
| Calculated Pouring Time | \( t \) | 25 | s |
| Choke Area (Calculated) | \( S_{\text{choke}} \) | 8.5 | cm² |
| Choke Area (Adopted) | \( S_{\text{choke}} \) | 12.0 | cm² |
| Metallostatic Head | \( H_p \) | 35 | cm |
Prior to implementing any feeding aids, a preliminary solidification simulation using MAGMA software was conducted. This step is crucial for nodular cast iron as its solidification behavior, with a significant graphite expansion phase, requires precise control of temperature gradients. The simulation clearly identified several major isolated thermal centers corresponding to the junctions of ribs, brackets, and the terminal box with the main barrel wall. The software also calculated the geometrical modulus \( M \) (Volume/Surface Area ratio) at these critical points, a key parameter for designing effective feeding. For example, the modulus at the upper mounting bracket hotspot was determined to be approximately 1.6 cm. The predicted shrinkage porosity locations from this initial simulation aligned perfectly with these high-modulus zones.
Based on the simulation feedback, a targeted feeding strategy was devised. Exothermic insulating sleeves (feeder heads) were selected and placed on the thermal centers located on the cope side, primarily the upper and lower mounting brackets. The selection criterion is that the feeder neck modulus must be greater than the casting hotspot modulus to ensure directional solidification towards the feeder. It is critical to note that placing a feeder actually increases the effective modulus of the casting section it contacts. Therefore, the feeder is selected for a modulus 20-30% larger than the initial simulated casting modulus. For the bracket with \( M_{\text{casting}} = 1.6 \, \text{cm} \), a feeder designed for a feeding modulus \( M_{\text{feeder}} \approx 1.9 – 2.1 \, \text{cm} \) would be appropriate. The relationship can be summarized as:
$$ M_{\text{feeder, required}} \geq (1.2 \text{ to } 1.3) \times M_{\text{casting, hotspot}} $$
Chills were applied to thicker sections on the drag side, such as certain internal rib intersections, to accelerate local solidification, eliminate isolated liquid pools, and promote a more favorable temperature gradient towards the applied feeders.
A second solidification simulation of this updated design, incorporating feeders and chills, showed a significant improvement. The temperature field demonstrated a clear directional solidification pattern, with the thermal centers now integrated into the feeder feeding paths. The final liquid pools were successfully displaced into the feeders, and the predicted shrinkage porosity in the main casting body was eliminated.
| Hotspot Location | Initial Modulus (cm) | Feeding Method | Purpose |
|---|---|---|---|
| Upper Mounting Bracket | 1.6 | Exothermic Insulating Feeder | Provide liquid feed metal to compensate for shrinkage |
| Lower Mounting Bracket | ~1.5 | Exothermic Insulating Feeder | Provide liquid feed metal to compensate for shrinkage |
| Internal Rib Junction (Drag) | ~1.4 | External Chill | Accelerate solidification, eliminate isolated liquid pool |
| Terminal Box Base | 1.3 | Initially none; later added Chill | Accelerate solidification to align with neighboring sections |
The first trial casting produced using the designed process was inspected via X-ray. While most areas met the specification, a localized shrinkage porosity defect was detected in the region where the terminal box connects to the barrel wall—an area classified as a non-critical zone but required to meet Class 3. Analysis revealed that this particular junction, though smaller than the main brackets, still formed a sufficient thermal center. In the initial design, it solidified slightly later than the surrounding barrel wall but was not adequately fed by the main feeders or effectively chilled, resulting in a small, isolated liquid pool that developed microshrinkage upon solidification.
The corrective action was precise and informed by the defect analysis. Instead of adding another feeder (which would reduce yield and could over-constrain the casting), two small, strategically placed chills were added directly to the mold cavity adjacent to the terminal box junction. The function of these chills is governed by the principle of extracting heat rapidly. The chill’s ability to absorb heat \( Q \) can be approximated by:
$$ Q = m_{\text{chill}} \cdot c_{\text{chill}} \cdot \Delta T_{\text{chill}} + m_{\text{chill}} \cdot L_{\text{fusion}} $$
where \( m_{\text{chill}} \) is the chill mass, \( c_{\text{chill}} \) is its specific heat capacity, \( \Delta T_{\text{chill}} \) is its temperature rise, and \( L_{\text{fusion}} \) is the latent heat of fusion if the chill material itself melts (e.g., in the case of internal chills). In this case, external cast iron chills were used. Their purpose was to increase the solidification rate of the terminal box hotspot, effectively reducing its local modulus and synchronizing its solidification time with the adjacent thinner sections. This eliminated the last-to-freeze isolated liquid pool.
A subsequent simulation confirmed the efficacy of this change, showing the terminal box region solidifying in unison with its surroundings. The second trial casting was produced with the optimized process including the terminal box chills. X-ray inspection confirmed the complete absence of shrinkage defects above the acceptable Class 3 level in all non-critical areas, with critical areas well within the stricter Class 2 requirement. The chemical composition and mechanical properties of the cast nodular cast iron housing were verified, meeting and exceeding the specifications for QT500-7, as detailed below.
| Element | Content (wt. %) | Element | Content (wt. %) |
|---|---|---|---|
| Carbon (C) | 3.64 | Sulfur (S) | 0.010 |
| Silicon (Si) | 2.83 | Copper (Cu) | 0.40 |
| Manganese (Mn) | 0.38 | Magnesium (Mg) | 0.045 |
| Phosphorus (P) | 0.029 |
| Property | Specification (QT500-7) | Measured Value |
|---|---|---|
| Tensile Strength | ≥ 500 MPa | 570 MPa |
| Yield Strength | ≥ 320 MPa | 423 MPa |
| Elongation | ≥ 7 % | 15.5 % |
| Hardness | 170 – 230 HBW | 200 HBW |
The successful development of this complex nodular cast iron motor housing underscores the critical importance of an integrated approach combining fundamental casting principles with advanced simulation tools. The modulus, a classical geometric calculation, remains a cornerstone for designing feeding systems, but modern simulation software allows for its rapid and accurate determination in complex geometries, bypassing tedious manual calculations. More importantly, simulation provides visual and quantitative insight into the dynamic solidification sequence, revealing not just the location of thermal centers but also the formation and movement of isolated liquid regions. A key learning point is that a low “shrinkage propensity” prediction in simulation does not guarantee a defect-free casting; one must critically examine the solidification progression for any potential isolated liquid pools, the formation of which can be sensitive to specific simulation parameters and actual production variables like precise chemistry and pouring temperature. The final process—a horizontal parting, a carefully calculated pressurized gating system, a combination of exothermic feeders for major hot spots, and strategic use of chills for secondary thermal centers—proved to be a robust and yield-efficient solution for producing sound, high-integrity thin-walled nodular cast iron castings.