We have been working on the production of thin-walled motor frames for submersible pumps. These frames, with a typical wall thickness of 11–12 mm, are classified as thin-walled castings. The material specification requires HT250 gray iron to ensure adequate strength and pressure tightness under a maximum water pressure of 0.2 MPa. However, during our routine production, we encountered severe quality issues concentrated on the frame flange—the critical sealing surface where the end cover is mounted. The defects included gas holes, shrinkage cavities, and sand holes (sand inclusion), all of which directly compromised the air-tightness after assembly. After systematic analysis, we identified that these sand casting defects originated from multiple interrelated factors: improper gating system design, suboptimal molten iron composition, inadequate molding sand control, excessive gas evolution during pouring, and poorly designed mold. In this article, we share our comprehensive approach to eliminating these defects, emphasizing the role of sand casting defect analysis and process optimization. We will present our findings using data tables and mathematical formulas to summarize the control parameters.
1. Introduction
Submersible pump motors operate in underwater environments and require hermetic sealing to prevent water ingress. The frame flange, which houses the O-ring seal, is the most demanding feature in terms of density and surface integrity. Any sand casting defect on this flange—whether it is a gas pore, shrinkage cavity, or sand inclusion—creates a leak path. Initially, our rejection rate for machined flanges was as high as 15%, with the majority of defects located on the cope side of the mold at the flange position. Through careful metallographic and visual inspection, we categorized the defects:
- Gas holes: spherical or elongated voids with smooth inner walls, often caused by trapped gas from molding sand or core binders.
- Shrinkage cavities: irregular, dendritic voids at hot spots, indicating insufficient feeding during solidification.
- Sand holes: irregular cavities containing loose sand particles, resulting from sand erosion or loose sand falling into the cavity.
All of these are classic manifestations of sand casting defect mechanisms. To systematically resolve them, we implemented improvements in five areas: pouring process, iron chemistry, molding sand properties, gas evolution control, and gating system design. Each aspect is discussed below with quantitative targets.
2. Analysis of Sand Casting Defect Root Causes
Before presenting the improvements, we performed a detailed root cause analysis. The primary sand casting defect modes and their contributors are summarized in Table 1.
| Defect Type | Appearance | Root Cause(s) |
|---|---|---|
| Gas hole (blowhole) | Smooth, spherical/elongated, often near surface | Excessive gas evolution from mold/core; low permeability; insufficient venting |
| Shrinkage cavity | Dendritic, rough internal surface, at hot spots | Inadequate feeding; high carbon equivalent; poor gating design |
| Sand hole (sand inclusion) | Irregular cavity with loose sand grains | Low mold surface hardness; sand erosion during pouring; loose sand from core or mold assembly |
All three defect types appeared predominantly on the cope side of the mold at the flange position, indicating that the defect formation was closely related to the gas evolution and thermal gradient in that region. The flange is a relatively thick section compared to the surrounding thin wall, acting as a hot spot that solidifies last. This makes it susceptible to shrinkage if the feed path is blocked, and also to gas entrapment because gas bubbles tend to migrate toward the last-solidifying region.
3. Process Improvements to Eliminate Sand Casting Defects
3.1 Pouring Process Control
To reduce the incidence of sand casting defects from mold erosion and loose sand, we enforced stricter molding practices. The surface hardness of the green sand mold was required to reach 60–70 kPa (measured by a mold hardness tester). We instructed operators to avoid scraping or damaging the mold surface when placing the core, and to carefully blow out any loose sand using compressed air before closing the mold. During pouring, we installed ceramic foam filters at the sprue base and in the gating system to prevent sand or slag from entering the cavity. Pouring speed was controlled to avoid turbulence: the initial pour was slow (about 0.5 m/s at the sprue) to allow gentle filling, followed by a steady rate. These measures significantly reduced the frequency of sand holes.
3.2 Control of Molten Iron Composition
The chemical composition of HT250 gray iron was optimized to minimize both gas-related and shrinkage-related sand casting defects. Carbon and silicon are strong graphitizers; excessively high levels coarsen the graphite and increase ferrite content, reducing strength and promoting shrinkage porosity. Conversely, too low carbon equivalent can cause chilling and hard spots. We conducted a series of trials to determine the optimal silicon-to-carbon ratio (Si/C). The results showed that maintaining the Si/C ratio between 0.60 and 0.70 provided the best balance of mechanical properties and casting soundness. Table 2 lists the final target composition range.
| Element | Target Range (wt%) | Effect on Sand Casting Defect |
|---|---|---|
| Total Carbon (C) | 3.20 – 3.40 | Higher C reduces shrinkage but increases gas solubility; controlled to avoid gas holes. |
| Silicon (Si) | 1.90 – 2.20 | Promotes graphite; Si/C ratio 0.60–0.70 reduces shrinkage. |
| Manganese (Mn) | 0.60 – 0.80 | Stabilizes pearlite; reduces gas porosity. |
| Phosphorus (P) | < 0.10 | High P increases brittleness and shrinkage tendency. |
| Sulfur (S) | 0.08 – 0.12 | Nucleates graphite; too high increases gas holes. |
The carbon equivalent (CE) is calculated using the standard equation:
$$ \text{CE} = \text{C} + \frac{\text{Si}}{3} + \frac{\text{P}}{3} $$
We targeted a CE range of 3.8 – 4.1%, which is typical for HT250. The Si/C ratio control was particularly effective in reducing shrinkage cavities. Figure 1 illustrates the relationship between Si/C ratio and defect rate observed in our trials.
3.3 Molding Sand Quality Control
Molding sand properties directly influence sand casting defects. We established strict specifications for the green sand system, as shown in Table 3.
| Property | Target Value | Rationale |
|---|---|---|
| Moisture content | 5.0 – 6.0% | Too high increases gas evolution; too low reduces mold strength. |
| Permeability | 50 – 100 (AFS units) | Ensures gases can escape through mold walls. |
| Green compression strength | 60 – 90 kPa | Prevents mold erosion and sand wash. |
| Sand grain size (AFS GFN) | 50 – 60 | Base sand 50/140 mesh, well distributed across 4 adjacent sieves. |
| Clay content | 8 – 12% | Bonding agent; excessive clay reduces permeability. |
The use of well-graded sand (with 10–35% retained on each of four consecutive sieves) minimized thermal expansion defects and ensured uniform compaction. The permeability was kept within the range to balance gas release and mold strength. We also implemented daily sand testing to maintain consistency.
3.4 Control of Gas Evolution and Venting
Gas holes, one of the most troublesome sand casting defects, were addressed by controlling both the gas evolution rate and the mold’s ability to vent. During pouring, the high-temperature metal vaporizes moisture and burns binders in the sand, generating gases. If the gas generation rate exceeds the venting capacity, the gas pressure at the mold-metal interface may exceed a critical value, causing bubbles to penetrate the liquid metal. We derived a simplified model for the critical interface pressure:
$$ P_{\text{crit}} = \frac{2\sigma}{r} $$
where \(\sigma\) is the surface tension of liquid iron (approximately 1.2 N/m at pouring temperature) and \(r\) is the radius of the bubble nucleus. For a typical bubble nucleus of 0.1 mm, the critical pressure is about 24 kPa. If the gas pressure in the mold exceeds this value, gas can be absorbed. To prevent this, we reduced the gas evolution rate by thoroughly drying the cores and the mold surface before pouring. A baking step at 150–200°C for 1–2 hours was added for all cores and the mold cavity surface (using a hot air blower). In addition, we increased the number of vent holes in the cope and core prints. The total vent area was calculated to ensure the actual gas pressure remains below the critical threshold. The gas evolution rate \(Q_g\) (cm³/s) from the sand was measured by a simple test, and the required vent area \(A_v\) was obtained from Darcy’s law for gas flow through sand:
$$ \frac{Q_g}{A_v} = \frac{k \Delta P}{\mu L} $$
where \(k\) is permeability (cm²), \(\Delta P\) is pressure drop (Pa), \(\mu\) is gas viscosity, and \(L\) is sand thickness. By designing adequate venting, we eliminated gas hole defects entirely.
3.5 Feeding System Redesign
Shrinkage cavities and remaining gas holes on the flange were finally solved by modifying the gating and risering system. The original design had the ingate at the bottom (bottom-gated) which provided smooth filling but created a temperature gradient that left the top flange as a hot spot with no effective riser. We added a blind riser on the flange (see Figure 1) and increased the ingate cross-section by 30% to promote directional solidification from the bottom upward. The riser volume was designed to compensate for the volumetric shrinkage of gray iron, which is about 1.5–2.0% for HT250. The shrinkage volume \(V_s\) is given by:
$$ V_s = V_c \cdot \varepsilon $$
where \(V_c\) is the casting volume and \(\varepsilon\) is the solidification shrinkage factor. For a typical frame of 5 kg, the shrinkage volume is about 80–100 cm³, which our riser (volume 120 cm³) easily accommodated. After the modification, radiography of the flange showed no shrinkage cavities.
4. Results and Discussion
We implemented all five improvements simultaneously on the production line and tracked the defect rate over a period of three months. The results are summarized in Table 4.
| Defect Type | Before Improvement (%) | After Improvement (%) | Reduction Factor |
|---|---|---|---|
| Gas holes | 6.2 | 0.8 | 7.8× |
| Shrinkage cavities | 4.5 | 0.3 | 15.0× |
| Sand holes | 4.3 | 0.5 | 8.6× |
| Total rejection rate | 15.0 | 1.6 | 9.4× |
The dramatic reduction in all three sand casting defect categories confirms the effectiveness of our integrated approach. Particularly, the elimination of gas holes was achieved through the combination of lower moisture, higher permeability, controlled Si/C ratio, and improved venting. The shrinkage cavities were virtually eliminated by the new riser design and optimized carbon equivalent. Sand holes were reduced by stricter mold handling and the use of filters. The total rejection rate dropped from 15% to 1.6%, leading to substantial cost savings and improved production throughput.
It is worth noting that many practitioners treat sand casting defect issues in isolation. Our study demonstrates that sand casting defects are often interlinked: for example, gas holes and sand holes can both be exacerbated by poor mold compaction. By addressing the process holistically—covering chemistry, sand properties, gating, and pouring—we achieved synergistic benefits. The lightweight thin-wall design (11–12 mm) is particularly sensitive to these defects; the solutions we developed are also applicable to similar thin-wall gray iron castings used in fluid-handling applications where pressure tightness is critical.
5. Conclusion
We have successfully resolved the sand casting defect problems on the flange of thin-walled submersible pump motor frames. Through systematic analysis and improvement of the pouring process, iron composition, molding sand quality, gas evolution control, and gating system design, the occurrence of gas holes, shrinkage cavities, and sand holes was reduced by more than 90%. The key lessons are:
- Maintaining the Si/C ratio between 0.60 and 0.70 effectively minimizes shrinkage porosity without sacrificing machinability.
- Controlling green sand moisture (5–6%) and permeability (50–100 AFS) together with adequate venting prevents gas hole formation.
- Adding a blind riser on the flange hot spot and enlarging the ingate promotes directional solidification and eliminates shrinkage cavities.
- Strict mold handling and ceramic foam filters eliminate sand holes.
Our approach provides a practical reference for foundries producing thin-wall pressure-tight gray iron castings, where any sand casting defect can lead to functional failure. By continuously monitoring these process parameters, we have maintained a defect rate below 2% and significantly improved the reliability of our submersible pump motors.