In our foundry, we recently encountered significant quality challenges during the production of a clutch housing casting for a major automotive client. This component is a classic example of a complex thin-wall iron casting with intricate internal cavities. The maximum envelope dimensions are 363 mm × 384 mm × 220 mm, the weight is 18 kg, and the material is HT250 (a grey cast iron with tensile strength ≥ 250 MPa). The smallest wall thickness is only 5 mm, and the part must pass an air-tightness test. These requirements demand excellent mechanical properties, a dense microstructure, and an absolute absence of internal casting defects. This case study details our systematic investigation into the root causes of three predominant sand casting defects—gas porosity, sand inclusion, and core breakage—and the comprehensive process improvements that ultimately resolved these issues. Throughout this discussion, we will repeatedly emphasize the term ‘sand casting defect’ as we analyze each failure mode and corrective action.
Our initial production process employed a green sand molding line with an impulse squeeze molding machine. The flask internal dimensions were 1200 mm × 800 mm × (350/350) mm. To maximize productivity and reduce cost, we adopted a one-mold-four-cavities layout. The gating system was designed as a mid-gating horizontal pour, which offers stable filling, reduced buoyant force on the cores, and good resistance to gas holes and slag inclusions. Each casting had three ingates placed on the parting line. All cores were made of resin-coated sand (hot-box process), which inherently generates a substantial volume of gas during pouring. To mitigate this, we initially incorporated a common exhaust channel for each pair of cores and placed open risers on the top flange, along with vent pins at all upper projections. Figure 1 in the original reference illustrated this layout. Unfortunately, this initial design was inadequate, leading to a high scrap rate due to three distinct sand casting defects.
After a batch of 500 castings, we performed a detailed statistical analysis of the defects. The results are summarized in Table 1, which clearly shows the localized nature of each type of sand casting defect.
| Defect Type | Location on Casting | Frequency (%) | Classification |
|---|---|---|---|
| Gas porosity | Bottom face highest point (root of vent slit) | 35 | Sand casting defect |
| Sand inclusion | Top flange area (near flask wall) | 40 | Sand casting defect |
| Core breakage | Three process-hole core prints connecting two cores | 25 | Sand casting defect (indirect) |
Each of these sand casting defects required a separate root-cause analysis before we could devise effective countermeasures.
Root Cause Analysis of Gas Porosity
Gas porosity is one of the most common sand casting defects in thin-wall iron castings. In our case, the porosity appeared exclusively at the highest point of the bottom face – exactly where the original vent slit was located. This pattern strongly suggested that the gas produced by the resin-coated sand cores could not escape quickly enough during pouring. The initial exhaust system provided only one common channel for each pair of cores, leading to a cross-sectional area that was too small to handle the combined gas evolution from two cores simultaneously. Moreover, the vent slit was positioned at the bottom of the casting, which is not the optimal location for gas removal; gases naturally rise.
We can quantify the required exhaust area using a simple volumetric gas evolution model. Let \(V_{core}\) be the volume of a single core, \(\rho_{sand}\) the density of the core sand, \(R_{gas}\) the gas yield per unit mass of binder, and \(t_{fill}\) the filling time. The total gas volume generated by two cores is:
$$ Q_{total} = 2 \cdot V_{core} \cdot \rho_{sand} \cdot R_{gas} $$
For our resin-coated sand, typical values are \(\rho_{sand} \approx 1.5 \, \text{g/cm}^3\) and \(R_{gas} \approx 15 \, \text{cm}^3/\text{g}\) at pouring temperature. With a core volume of approximately 800 cm³ for each core, and a filling time of about 4 seconds, the required exhaust area \(A_{exh}\) can be estimated from the Darcy flow equation for gases through sand:
$$ A_{exh} \geq \frac{Q_{total}}{v_{gas} \cdot \phi} $$
where \(v_{gas}\) is the permissible gas velocity (typically 1–2 m/s) and \(\phi\) is a permeability factor. Our original single-channel area was only 80 mm², which was insufficient to prevent backpressure that forced gas into the solidifying metal, forming gas pores. This defect is a classic sand casting defect driven by inadequate venting. To mitigate this, we decided to provide each core with its own dedicated exhaust channel, and to position the vent at the top of each core rather than sharing a common path. This simple change increased the total exhaust area to 160 mm² and eliminated the interference between adjacent cavities.
Root Cause Analysis of Sand Inclusion
Sand inclusion appeared predominantly on the top flange area, which was formed by the green sand mold. This area is close to the flask wall, where the impulse squeeze process often yields lower mold hardness due to the “bridging” effect at deep pockets. We measured the mold hardness at this vertical wall and found it to be only 50–60 on the mold hardness scale, far below the required 70. As a result, the turbulent iron flow during pouring eroded the weak sand surface, creating loose sand grains that became entrapped in the casting – a typical sand casting defect. Compounding the problem, the top flange is a thick section, which prolongs the liquid state and allows more time for sand erosion.
Correcting the low mold hardness would have required costly modifications to the pattern or the molding machine settings. Instead, we opted to change the core design. By increasing the size of core #2 (the upper core) so that it now forms the entire vertical face of the top flange, we replaced the weak green sand surface with a hard cured resin-coated sand surface. The core hardness is inherently above 90, eliminating the erosion problem. This approach directly addresses the root cause of this sand casting defect without altering the molding process. The modified cross-section is shown schematically in Figure 4 of the original reference. A comparison of the original and modified core prints is summarized in Table 2.
| Parameter | Original Design | Modified Design |
|---|---|---|
| Top flange forming surface | Green sand mold | Core #2 surface |
| Mold/core hardness at flange face | 50–60 | 90+ |
| Susceptibility to sand inclusion | High | Very low |
Root Cause Analysis of Core Breakage
The third sand casting defect we encountered was core breakage, specifically the fracture of male core prints at three locations where two separate cores (core #1 and core #2) were joined. These cores formed deep internal cavities and were essentially cantilevered; they relied on three small process-hole core prints to connect them into a single rigid assembly. The print diameter was only 10 mm, and the length was 15 mm. During core assembly, handling, and pouring, the bending moment at these prints exceeded the strength of the resin-coated sand, causing the male prints to snap. This led to misalignment, core shift, and ultimately scrapped castings with thin walls or metal penetration.
We calculated the approximate bending stress \(\sigma_{bend}\) at the print root using a simple cantilever beam model:
$$ \sigma_{bend} = \frac{32 M}{\pi d^3} $$
where \(M\) is the bending moment due to the weight of the core and buoyancy forces, and \(d\) is the print diameter. For a core mass of about 2.5 kg acting at a lever arm of 100 mm, \(M \approx 2.45 \, \text{N·m}\). With \(d = 10 \, \text{mm}\), the stress was approximately 25 MPa, which exceeded the flexural strength of our standard resin-coated sand (typically 20 MPa).
Rather than upgrading to a more expensive high-strength resin system (which would also increase gas evolution), we introduced a steel core rod reinforcement at these critical prints. We modified the core box for core #2 to create a 10 mm diameter hole that could accommodate a 10 mm steel rod. During core shooting, the rod was inserted into the hole and bonded by the resin. Correspondingly, the core box for core #1 was modified to produce a 12 mm diameter recess. Before assembly, we applied a small amount of core adhesive into the recess, then inserted the steel rod protruding from core #2 into the recess. The adhesive further strengthened the joint. This solution increased the effective bending strength of the connection by a factor of over 3, eliminating all core breakage defects. The reinforcement scheme is illustrated in Figure 5 of the original article. The stress reduction can be expressed as:
With the steel rod of diameter \(d_{rod} = 10 \, \text{mm}\) and steel modulus \(E_s \approx 200 \, \text{GPa}\) versus sand modulus \(E_sand \approx 10 \, \text{GPa}\), the composite section’s moment of inertia increases dramatically, reducing the stress in the sand matrix to safe levels.
$$ \sigma_{sand,new} = \frac{M y_{sand}}{I_{composite}} \ll 20 \, \text{MPa} $$
Process Improvement Implementation and Results
We implemented three process improvements simultaneously in a trial run of 200 castings. The changes are summarized in Table 3.
| Defect Addressed | Change Description | Key Parameter |
|---|---|---|
| Gas porosity | Dedicated exhaust channel for each core; relocated vents to highest point of core | Exhaust area increased from 80 mm² to 160 mm² |
| Sand inclusion (top flange) | Enlarged core #2 to form the flange vertical surface; removed deep pocket in green sand mold | Mold hardness replaced by core hardness >90 |
| Core breakage | Inserted steel rod (10 mm φ) in male prints; added adhesive assembly | Bending strength increased >3× |
After the trial, the defect rates dropped dramatically. Gas porosity scrap fell from 35% to under 2%. Sand inclusion defects completely disappeared. Core breakage was eliminated entirely. The overall casting yield improved from approximately 50% to 92%. The air-tightness test pass rate reached 100% for the remaining castings, confirming that the internal density was also satisfactory. Table 4 compares the before and after defect statistics.
| Defect Type | Before Improvement (%) | After Improvement (%) |
|---|---|---|
| Gas porosity | 35 | 1.5 |
| Sand inclusion | 40 | 0 |
| Core breakage | 25 | 0 |
| Overall scrap | 50 | 8 |
Beyond the immediate quality gains, the cost savings were substantial. The one-mold-four-cavities layout allowed us to maintain high productivity without sacrificing quality. The use of steel core rods added a minimal material cost (about $0.02 per casting), but this was far outweighed by the reduction in scrap. We also observed that the improved exhaust system reduced the incidence of other subtle sand casting defects like pinholes and subsurface blowholes, which had previously contributed to leakage failures during pressure testing. The elimination of sand inclusion on the flange also improved the surface finish, reducing subsequent machining issues.
Discussion and General Recommendations
This case study highlights how a systematic approach to identifying and eliminating sand casting defects can transform the production of complex thin-wall iron castings. The key lessons learned are:
- Gas-related defects are often a result of inadequate venting, especially when multiple cores are involved. Each core should have its own exhaust path, and vents must be placed at the highest points of the core to take advantage of natural buoyancy. The cross-sectional area of vent channels should be calculated based on the gas evolution rate, which can be estimated from core volume, binder content, and filling time.
- Sand inclusion defects can be avoided by ensuring that all surfaces exposed to high-velocity metal flow are formed by high-strength cores rather than low-hardness green sand. Whenever possible, deep vertical walls near the flask should be cored out to prevent erosion.
- Core breakage in slender prints can be mitigated by embedding steel reinforcements. This is often more economical than upgrading the core sand system and avoids increasing gas evolution. The adhesive assembly step adds minimal cycle time but greatly improves joint integrity.
- The phrase ‘sand casting defect’ encompasses a wide range of phenomena, but the root causes are often interrelated. For example, poor venting can lead to gas porosity, but also to increased backpressure that may cause core shift or even breakage. Therefore, a holistic view of the entire mold filling and solidification process is essential.
Other foundries facing similar sand casting defects in clutch housings or comparable complex thin-wall castings can adopt these principles. We recommend performing a detailed defect mapping and statistical analysis before making changes, as the localization often points directly to the root cause. Simple calculations using the formulas provided above can guide the sizing of vents and reinforcements.
In conclusion, our experience demonstrates that with careful analysis and targeted modifications, even severe sand casting defects can be eliminated without major capital investment. The one-mold-four-cavities process now runs reliably with a yield above 90%, producing clutch housings that consistently meet the air-tightness and mechanical property requirements. This success has become a benchmark for similar components in our production portfolio.