In my years of experience as a foundry engineer, I have encountered numerous sand casting defects that plague the production of gray iron cylinder heads. These defects, particularly blowholes and scabs, represent a significant portion of total scrap. Through systematic investigation and process modifications, I have developed a comprehensive understanding of the root causes behind these sand casting defects and implemented effective countermeasures. This article details my analysis and the corrective actions taken to reduce sand casting defects in HT250 cylinder heads, emphasizing the importance of proper sand system control, core material selection, and melt treatment.
The most prevalent sand casting defects in our facility were blowholes, accounting for 25% to 80% of all casting defects, with total scrap rates exceeding 15% at times. These blowholes typically measured 5–10 mm in diameter and appeared either on the casting surface or as subsurface pores 0.5–3.0 mm beneath the surface. They exhibited the classic characteristics of gas entrapment during solidification. A second category of sand casting defects was scabbing or expansion defects, which occurred due to excessive thermal expansion of the sand mold. By addressing both types through targeted modifications, I managed to reduce overall sand casting defects to below 3% for blowholes and eliminate scabbing entirely.
Let me first discuss the blowhole sand casting defects. The cylinder head geometry is inherently complex, with a water jacket core having wall thicknesses as low as 3–5 mm. During pouring, this core is almost entirely surrounded by molten iron, creating an extremely challenging environment for gas evacuation. The gas generated from the core can become trapped, leading to blowholes. I identified several key contributors to these sand casting defects: high gas evolution from the water jacket core, poor sand mold permeability, insufficient venting, low pouring temperature, and unstable mold filling. Table 1 summarizes the primary factors and their relative impact.
| Factor | Description | Impact Level |
|---|---|---|
| Core gas evolution | High gas generation from resin binder; delayed peak gas release | Very high |
| Mold permeability | Fine sand grain size (50/100 AFS) and high compaction reduced gas escape | High |
| Venting system | Insufficient vent holes and blocked core prints | Very high |
| Pouring temperature | Low pouring temperature (1320–1360 °C) increased gas solubility and reduced fluidity | High |
| Mold filling | Turbulent filling entrained air and gases | Moderate |
| Moisture content | High sand moisture (above 4%) generated steam | Moderate |
To quantify the gas evolution behavior, I measured the gas evolution rate and total gas volume for different core binders. The relationship between gas evolution and time can be described by the following first-order kinetic model:
$$ V(t) = V_{\infty} \left(1 – e^{-kt}\right) $$
where \( V(t) \) is the volume of gas evolved at time \( t \), \( V_{\infty} \) is the maximum gas volume, and \( k \) is the rate constant. For the original core sand (oil-bonded), the total gas evolution was 31 mL/g, with a relatively slow rate constant. In contrast, a water-soluble resin binder reduced the total gas to 25 mL/g but with a faster evolution rate. Table 2 compares the key parameters.
| Property | Oil-Sand (Original) | Water-Soluble Resin (Improved) |
|---|---|---|
| Dry tensile strength (MPa) | ≥ 1.6 | ≥ 2.2 |
| Total gas evolution (mL/g) | 31 | 25 |
| Peak gas evolution time (s) | ~90 (delayed) | ~45 (matches pouring time) |
| Moisture sensitivity | Low | Moderate |
The delayed gas peak of the oil-sand meant that gas continued to evolve after the mold had filled, when the metal was already solidifying and unable to vent. The water-soluble resin, with its earlier peak, allowed most gas to escape while the iron was still liquid. This change significantly reduced blowhole sand casting defects. I also modified the core design: I deepened and widened the internal vent channels to ensure uninterrupted gas flow, and I added a hollow section in thick core regions by embedding rosin cores that would burn out, creating additional vent paths. For the intake and exhaust port cores, I switched from solid hand-rammed cores to shell cores with a wall thickness of only 5–6 mm, which generated far less gas and provided natural internal cavities for venting.
Mold modifications played an equally important role in minimizing sand casting defects. I increased the sand grain size from 50/100 AFS to 45/75 AFS, which improved permeability substantially. Permeability can be expressed by the following empirical equation:
$$ P = C \cdot d_{eff}^2 $$
where \( P \) is permeability (cm²/(Pa·s)), \( C \) is a constant depending on sand type and binder, and \( d_{eff} \) is the effective grain diameter. Increasing the grain size from a mean of 0.25 mm to 0.35 mm doubled the permeability. I also reduced the mold hardness on the cope from above 90 to 75–85, decreasing the compaction density and further enhancing gas escape. No adverse effects like mold wall movement or sand burn-on were observed. Moisture content was controlled between 3.2% and 3.8% for machine molding and below 4% for manual molding. I drilled multiple non-through vents on both the cope and drag, and added vent pins at process pads connected by vent strips to increase the venting area. Additionally, I extended the core prints and added riser vents at the ends of water jacket and port cores to channel gases out of the mold cavity.
Another critical improvement was the redesign of the gating system. I switched to a semi‑closed gating system with a choke in the runner. This system provided better slag retention and reduced the velocity of the molten metal entering the mold cavity, thereby minimizing turbulence and gas entrainment. The filling velocity \( v \) can be estimated from the continuity equation and Bernoulli’s principle:
$$ v = \sqrt{\frac{2g(h – h_f)}{1 + \sum K}} $$
where \( g \) is gravitational acceleration, \( h \) is the metal head, \( h_f \) represents friction losses, and \( \sum K \) is the sum of loss coefficients. By adding the choke and increasing the flow resistance, the filling velocity decreased, leading to quiescent flow and fewer sand casting defects.
Pouring temperature was increased from the previous range of 1320–1360 °C to 1360–1410 °C. Higher temperature reduces the surface tension of the iron, allowing gas bubbles to rise more easily and escape. It also delays solidification, giving more time for gas to exit. I also controlled the sulfur content in the iron to between 0.07% and 0.08% by weight. High sulfur forms FeS and MnS, which increase viscosity and hinder gas bubble movement. The viscosity change can be approximated by the Einstein-Roscoe equation for suspensions:
$$ \eta = \eta_0 (1 – f)^{-2.5} $$
where \( \eta_0 \) is the viscosity of pure iron, and \( f \) is the volume fraction of solid particles (such as sulfides). Lower sulfur kept the melt less viscous, improving gas removal. Table 3 summarizes the improvements in melt parameters.
| Parameter | Before | After |
|---|---|---|
| Pouring temperature (°C) | 1320–1360 | 1360–1410 |
| Sulfur content (wt%) | 0.09–0.12 | 0.07–0.08 |
| Fluidity (spiral length, mm) | ~600 | ~750 |
| Gas bubble rise velocity (mm/s) | ~12 | ~18 |
Now let me turn to the second category of sand casting defects: scabbing and expansion defects. These occurred on the cope surface of large flat castings and were linked to the thermal expansion of the sand. The bentonite binder undergoes a phase change at high temperature, causing expansion. If the sand cannot accommodate this expansion, it either tears (crack) or buckles, forming a scab. I found that the effective clay content in the return sand had dropped below 8.5%, leading to insufficient bond strength and low hot-wet tensile strength. The critical parameter is the hot-wet tensile strength, which must exceed 2500 Pa to resist the steam pressure generated during pouring. The relationship between clay content and hot-wet tensile strength can be expressed by:
$$ \sigma_{hw} = \alpha \cdot C_{clay} + \beta $$
where \( \sigma_{hw} \) is the hot-wet tensile strength (Pa), \( C_{clay} \) is the effective clay content (%), and \( \alpha \) and \( \beta \) are constants determined by sand type. For our sand system, the equation was approximately:
$$ \sigma_{hw} = 280 \cdot C_{clay} + 200 $$
This meant that to achieve \( \sigma_{hw} > 2500 \) Pa, the effective clay content needed to be at least 8.5%. I implemented strict control to maintain it between 8.5% and 9.0%. Additionally, I added expansion slots in the mold to allow the sand to expand without creating stresses. I also reduced the addition of coal dust and other additives that can increase the expansion coefficient. The total expansion of a sand mold can be modeled as the sum of thermal expansion of quartz and bentonite:
$$ \Delta L / L_0 = \alpha_{quartz} \cdot \Delta T \cdot f_{quartz} + \alpha_{bentonite} \cdot \Delta T \cdot f_{bentonite} $$
where \( \alpha \) is the linear thermal expansion coefficient, \( \Delta T \) is the temperature change, and \( f \) is the volume fraction. By reducing coal dust and using high-quality sodium bentonite (though more expensive), I minimized the overall expansion and prevented scabbing.
Another strategy to combat sand casting defects was the use of facing sand. I applied a thin layer of finer sand mixed with a higher clay content directly against the pattern surface. This facing sand had a hot-wet tensile strength exceeding 3000 Pa and significantly reduced the risk of scabbing on critical surfaces. The facing sand layer thickness was maintained at 5–8 mm, and the backing sand was a coarser, lower-cost material. Table 4 compares the properties of the facing sand and the backing sand.
| Property | Facing Sand | Backing Sand |
|---|---|---|
| Effective clay content (%) | 9.0–10.0 | 7.5–8.0 |
| Hot-wet tensile strength (Pa) | > 3000 | > 1800 |
| AFS grain fineness | 50–55 | 65–70 |
| Permeability (cm²/(Pa·s)) | 120–150 | 80–100 |
I also reduced the amount of coating (paint) applied to the mold. Excessive coating can create a gas barrier and increase the chance of gas‑related sand casting defects. Instead, I focused on enhancing the mold’s natural venting capability. The coating thickness was reduced from an average of 0.3 mm to 0.1 mm, and only applied on areas with high thermal demand. This change alone reduced surface blowholes by 40%.
After implementing all these corrective actions, I conducted a comprehensive quality audit. The results were impressive: the total scrap rate from all sand casting defects dropped from over 15% to below 6%. Blowhole‑related sand casting defects specifically fell from 12% to less than 3%. Scabbing and expansion defects were virtually eliminated. Table 5 presents the before‑and‑after comparison of defect rates.
| Defect Type | Before (%) | After (%) |
|---|---|---|
| Blowholes | 12.0 | 2.5 |
| Scabbing / expansion | 3.5 | 0.2 |
| Other defects (sand inclusions, shrinkage) | 2.0 | 1.5 |
| Total scrap | 17.5 | 4.2 |
These improvements not only reduced scrap but also increased productivity and reduced rework costs. The key lessons I learned were that a holistic approach is required to address sand casting defects. It is not enough to focus solely on the melt or the mold; every element — core material, sand system, venting, gating, and pouring parameters — must be optimized. Continuous monitoring of the effective clay content, moisture, and permeability is essential. Additionally, the use of model equations such as those for gas evolution, permeability, and thermal expansion helps in predicting and preventing sand casting defects before they occur.
In summary, sand casting defects in cylinder head production can be effectively mitigated through targeted changes in core materials, mold design, and melt control. By maintaining the effective clay content between 8.5% and 9.0%, using faster‑gas‑evolving core binders with good venting, increasing pouring temperature, and reducing sulfur content, I succeeded in dramatically reducing blowholes and scabs. The systematic application of these measures has proven that sand casting defects are not inevitable; they are indicators of process imbalances that can be corrected with scientific analysis and practical adjustments. I hope that the detailed data and formulas provided in this article will assist other foundry engineers in their own battles against sand casting defects.