In our foundry, the WP12 cylinder block is one of the most complex and mechanically demanding components for high-speed diesel engines. Each rough casting weighs 310 kg and is produced on a KW molding line with green sand process. The core package consists of 24 sand cores, including six large cylinder cores, front and rear end cores, water jacket cores, tappet core assemblies, and top cover cores. The main core set is assembled, dipped in coating, dried in a surface drying oven, and then lowered into the lower mold which has been sprayed with a mold coating enhancer. The pouring temperature is controlled between 1385°C and 1395°C with a pouring time of around 30 seconds. Despite careful process design, a severe sand casting defect known as “peeling and sand inclusion” (often referred to as scabbing or sand expansion defect) has plagued our production, accounting for over 60% of total scrap at peak times. This article presents our systematic investigation and countermeasures that successfully reduced the scrap rate from 10.9% to 2.1%.
Defect Characteristics and Statistical Data
The typical sand casting defect appears as a thin layer of metal with embedded sand particles on the casting surface, either on the side or bottom of the cylinder block. This defect occurs when the sand mold surface undergoes rapid thermal expansion and moisture migration during pouring, causing the surface layer to delaminate, buckle, and peel off. The hot metal then fills the gap, creating a sand inclusion. The defect distribution was recorded over two months as shown in Table 1.
| Month | Total Castings Produced | Scrap Due to Sand Casting Defect | Scrap Rate (%) |
|---|---|---|---|
| September | 3492 | 456 | 13.06 |
| October | 5174 | 492 | 9.50 |
The data clearly indicate that this sand casting defect was intolerable and demanded immediate action. A cross-functional team was formed to track every step of the production flow and analyze the root causes.
Root Cause Analysis
Through detailed observation and data collection, we identified several contributory factors that lead to the formation of this particular sand casting defect:
- Hot sand issue: The average temperature of return sand exceeded 55°C, causing large fluctuations in mold moisture content.
- Unstable return sand properties: Two molding lines shared the same sand system, but the sand properties varied significantly, deviating from the required specifications.
- Aged dust collection equipment: The efficiency of the dust removal system was inconsistent, failing to effectively remove dead fines from the sand, which degraded molding sand performance.
- Fluctuating quality of raw materials: The bentonite came from multiple suppliers with variable quality, directly affecting the hot-wet tensile strength of the sand.
- High temperature of core set when placed into mold: The main core set after drying could reach 50°C, which preheated and dried out the lower mold surface, creating a weak layer prone to scabbing.
- Improper gating system design: The original middle-injection system caused uneven filling and excessive thermal shock to certain mold areas.
- Structural features of the cylinder block: Large flat surfaces on the casting generated high thermal expansion stresses during pouring, exacerbating the sand casting defect.
Countermeasures Implemented
1. Optimization of Molding Sand Properties
The first line of defense against any sand casting defect is proper control of the molding sand. We systematically adjusted the sand recipe based on statistical analysis of return sand characteristics. The target composition for return sand is shown in Table 2.
| Parameter | Target Range |
|---|---|
| Bentonite content (%) | 6.8 – 7.8 |
| Coal dust content (%) | 4.8 – 5.6 |
| Clay content (%) | 11 – 13 |
| AFS grain fineness (central fraction) (%) | ≥ 70 |
| Dead fines content (%) | < 2.5 |
| Pre-mixing water spray required | Yes |
For the single sand (the common sand used on the molding line), we specified the properties listed in Table 3.
| Parameter | Target Value |
|---|---|
| Wet compressive strength (×10⁵ Pa) | 1.60 – 1.80 |
| Bentonite addition (kg per batch) | 19 – 30 |
| Moisture content (%) | 3.0 – 3.4 |
| Compactability (%) | 39 – 43 |
| Hot-wet tensile strength (×10² Pa) | ≥ 40 |
| Permeability | 100 – 130 |
The hot-wet tensile strength is a critical parameter that directly indicates the resistance of the sand to the sand casting defect. The relationship between the hot-wet tensile strength ($\sigma_{hw}$) and the moisture content ($w$) as well as the bentonite quality can be approximated by:
$$ \sigma_{hw} = k \cdot \frac{B}{w^{0.5}} \cdot (1 – e^{-\alpha t}) $$
where $B$ is the effective bentonite binder content, $w$ is the moisture percentage, $t$ is the heating time, $k$ and $\alpha$ are empirical constants. Improving the bentonite quality (higher smectite content) directly increases $B$ and thus $\sigma_{hw}$, reducing the likelihood of the sand casting defect.
In addition, we introduced a separate facing sand for the critical areas of the mold. The facing sand was prepared with higher hot-wet tensile strength and lower permeability to provide a tougher surface layer. Table 4 lists the target properties of the facing sand.
| Parameter | Target Value |
|---|---|
| Moisture content (%) | 4.2 – 5.0 |
| Wet compressive strength (×10⁵ Pa) | ≥ 1.45 |
| Permeability | ≥ 80 |
| AFS grain fineness | 45 – 55 |
| Hot-wet tensile strength (×10² Pa) | ≥ 45 |
The facing sand was applied manually onto the pattern plate at the exact locations where the sand casting defect frequently occurred. This simple but effective measure significantly reduced the incidence of scabbing.
2. Control of Core Quality and Temperature
Since the WP12 cylinder block uses 24 cores, any variation in core quality can trigger a sand casting defect. We established strict process parameters for core coating density, drying temperature, and residual moisture content. Table 5 summarizes the key parameters.
| Core Type | Drying Temperature (°C) | Dwell Time (min) | Coating Density (g/ml) | Max Residual Moisture (%) |
|---|---|---|---|---|
| Water jacket core | 180 – 210 | 10 | 1.35 – 1.40 | ≤ 0.3 |
| Tappet core assembly | 180 – 210 | 15 | 1.35 – 1.40 | ≤ 0.4 |
| Main cylinder core set | 180 – 210 | 25 | 1.15 – 1.20 | ≤ 0.5 |
Of particular importance was the temperature of the main core set when it was lowered into the lower mold. Initially, the core temperature could reach 50°C, which dried out the adjacent sand surface and promoted moisture migration. The moisture condensation zone created a weak layer that easily spalled, leading to the sand casting defect. We installed forced air cooling stations along the core assembly line to bring the core temperature below 40°C. The relationship between the critical time to scabbing ($t_{scab}$) and the temperature difference between core and sand ($\Delta T$) can be approximated by:
$$ t_{scab} \propto \frac{1}{\Delta T^2} $$
Even a small reduction in $\Delta T$ significantly delays the onset of the defect.
3. Gating System Redesign
The original middle-gate system introduced iron through the cylinder core prints at the bottom of the block, with ingates located at the bearing saddles. Although this avoided direct erosion of the mold surface, the filling was not smooth and gases could not escape easily. We redesigned the system to incorporate bottom-gating, which ensured a calm, progressive filling of the mold cavity and rapidly covered the entire lower mold surface with molten metal, reducing localized overheating. The position of the bottom ingates was also shifted away from the areas most prone to the sand casting defect. Additionally, we enlarged the risers and venting channels to reduce the pressure of hot gases on the mold surface. The modified gating system is described by the following flow rate equation for a bottom-gated system:
$$ Q = A \sqrt{2 g h} $$
where $Q$ is the flow rate, $A$ is the total ingate area, $g$ is gravitational acceleration, and $h$ is the effective metal head. By increasing $A$ and adjusting $h$, we achieved a more controlled filling rate and minimized thermal shock.
4. Structural Modification and Anti-Scab Nails
The casting geometry itself contributed to the sand casting defect. The large flat surfaces on the side of the cylinder block created a wide expanse of sand that expanded rapidly under heat. We introduced reinforcing ribs at those locations (without compromising the functional requirements) to break up the large flat area into smaller segments, reducing the thermal stress. However, the defect persisted on the adjacent curved surfaces. To address this, we designed and applied simple anti-scab nails (metal pins) on the pattern plate. These nails, after molding, created small recesses at the mold surface. During pouring, the molten iron fills these recesses, effectively pinning the sand layer in place and increasing the hot-wet tensile strength locally. The effect can be modeled as an increase in effective binder strength in the pinned area:
$$ \sigma_{effective} = \sigma_{sand} + \frac{F_{nail}}{A_{nail}} $$
where $F_{nail}$ is the restraining force provided by the nail and $A_{nail}$ is the area of influence per nail. This approach dramatically reduced the sand casting defect on the curved side surfaces.
Figure above illustrates typical manifestations of the sand casting defect that we targeted.
Results and Discussion
After implementing the above countermeasures, we tracked the scrap rate for the following months. Table 6 shows the dramatic reduction.
| Time Period | Total Castings Produced | Scrap Due to Sand Casting Defect | Scrap Rate (%) |
|---|---|---|---|
| November (optimization phase) | 4800 | 288 | 6.00 |
| December (final stabilization) | 5200 | 109 | 2.10 |
The scrap rate attributable to the sand casting defect dropped from a peak of 13.06% to 2.1%, representing a reduction of over 84%. This improvement has been sustained in subsequent production runs. The combination of sand control, core temperature management, gating redesign, and structural reinforcement proved to be a robust solution.
Conclusion
The sand casting defect known as peeling and sand inclusion in WP12 cylinder blocks was successfully mitigated through a systematic, multi-faceted approach. Key measures included optimizing the molding sand composition to improve hot-wet tensile strength, reducing core temperature before assembly, redesigning the gating system for smoother filling, and using anti-scab nails on critical mold surfaces. Each intervention addressed a specific root cause, and the cumulative effect yielded a stable, low-defect production process. This case demonstrates that even complex sand casting defect issues can be resolved by careful data collection, targeted experimentation, and disciplined process control.