In our foundry, we produce gray iron cylinder blocks for automotive engines. The cylinder block is a complex component with water jacket cavities, oil passages, and thin walls of 4 mm. We use a high-pressure molding line with cold-box cores and a hot-box water jacket core. Initially, the combined internal and external scrap rate reached 18%, with defects such as blowholes, double skin, sand inclusions, cutter damage on machined surfaces, oil inlet cracks, and mold collapse. Over time, we systematically investigated each type of sand casting defect and implemented targeted countermeasures. This article shares our experience in identifying root causes and eliminating these defects, using tables and formulas to summarize key relationships.
1. Blowhole Defects
Blowholes appeared on the water pump face, oil gallery face, stamping surface, glue spots, and water jacket out-sand holes. The root cause was always inadequate venting or gas entrapment. We derived a fundamental relation for gas pressure inside a core cavity during pouring:
$$ P_{\text{gas}} = \frac{nRT}{V} + \rho g h $$
where \(n\) is moles of gas produced by the core binder, \(R\) is the gas constant, \(T\) is temperature, \(V\) is the vent volume, \(\rho\) is molten iron density, and \(h\) is the height difference. To avoid blowholes, we must ensure \(P_{\text{gas}} < P_{\text{metal}} + P_{\text{vent resistance}}\). Table 1 summarizes each blowhole type and our countermeasures.
| Blowhole Location | Root Cause | Countermeasure |
|---|---|---|
| Water pump face | Excessive gas from thick water jacket core; metal backflow into vent | Replace open vents with blind vents; reduce vent pin height from 185 mm to 165 mm; add exhaust fins |
| Stamping surface | Last area to fill; insufficient venting | Add an additional overflow riser; increase riser contact area |
| Glue spots | Excess adhesive volatilizing in narrow spaces | Relocate glue spots to larger, more robust areas; reduce glue quantity |
| Water jacket out-sand holes | Metal backflow into core vent increases gas load | Eliminate drilling of vents in water jacket core; shorten adjacent vent pins |
After implementing these changes, the blowhole scrap rate dropped from 0.45% to below 0.015% for the water pump face, and similarly for other locations. This demonstrates that proper vent design is crucial in controlling this sand casting defect.
2. Double Skin Defect
Double skin occurred on the upper oil gallery side, a 4 mm thick wall. The thin section was the last to fill, and gas trapped beneath the iron caused incomplete fusion. Using the solidification theory, the critical condition for avoiding double skin can be expressed as:
$$ \frac{V_{\text{vent}}}{A_{\text{interface}}} \geq \frac{v_{\text{gas}}}{v_{\text{metal}}} \cdot t_{\text{solidification}} $$
where \(V_{\text{vent}}\) is vent volume, \(A_{\text{interface}}\) is the area of the thin wall, \(v_{\text{gas}}\) is gas evolution rate, \(v_{\text{metal}}\) is metal filling rate, and \(t_{\text{solidification}}\) is local solidification time. We added two blind vent pins directly at the double skin location (the opposite side was a 15 mm groove, preventing a pressure pad). The defect disappeared immediately.
| Defect Location | Cause | Measure |
|---|---|---|
| Upper oil gallery side | Gas trapped under thin wall; insufficient venting | Add two blind vent pins |
3. Sand Inclusion Defects
Sand inclusions were the most common sand casting defect, appearing on oil inlets, front face, stamping surface, bearing seats, cylinder bores, and the head face. We classified them by source: core sand, molding sand, coating flakes, and loose sand from tools. Table 2 lists each type.
| Location | Source of Sand | Root Cause | Countermeasure |
|---|---|---|---|
| Oil inlet | Core sand from crankcase core nozzle | Knocking off loose sand during core finishing | Prohibit hammering; use file instead; blow off all loose sand with air gun |
| Front face | Core sand from side core positioning slot | Robot gripper damaged the slot, falling into cavity | Adjust gripper claws; avoid contact with positioning slots |
| Stamping surface | Molding sand from crush ribs near overflow riser | Too deep crush rib (1 mm) crushed the sand | Reduce crush rib depth to 0.5 mm |
| Bearing seat | Vent pin loose sand | Scraped sand on top of vent pins fell into cavity | Shorten six vent pins that were unnecessary; eliminate manual poking |
| Cylinder bore (core sand) | Sand from water jacket core falling during handling | Core assembly protruding beyond pallet edge | Add guard rails to the core pallet |
| Cylinder bore (coating lumps) | Undissolved coating lumps | Insufficient stirring and filtering | Install automatic stirrer; clean coating tank regularly |
| Head face (coating flakes) | Detached coating due to loose water jacket core | Weak glue spots caused core movement | Relocate glue spots to larger areas (same as blowhole fix) |
We also developed a visual method to distinguish core sand from molding sand: core sand particles have sharp edges and a clean surface, while molding sand particles are rounded and covered with dark binder.
4. Machining Surface Cutter Damage
During machining of the reference surface, cutters were frequently broken by large burrs. The burrs were originally oriented parallel to the head face, so the grinding machine could not remove them. We changed the burr orientation to be perpendicular to the head face, allowing the grinding wheel to cut them off completely. Table 3 summarizes the change.
| Defect | Root Cause | Countermeasure |
|---|---|---|
| Cutter damage on reference surface | Burrs parallel to head face, not removed by grinder | Redesign burr orientation to perpendicular; use grinding machine |
5. Oil Inlet Cracks
During machining, we observed 2–3 cracked cylinder blocks per shift at the oil inlet. Investigation revealed that the support pad of the deburring machine was located exactly under the oil inlet. The downward force during deburring created excessive stress. Using stress concentration formula:
$$ \sigma_{\text{max}} = K_t \cdot \frac{F}{A} $$
where \(K_t\) is the stress concentration factor at the oil inlet geometry, \(F\) is the applied force, and \(A\) is the cross-sectional area. By cutting away a portion of the support pad directly under the oil inlet, the force was redistributed, eliminating the crack. Table 4 shows the fix.
| Defect | Root Cause | Countermeasure |
|---|---|---|
| Oil inlet crack after machining | Support pad concentrated force at oil inlet | Cut away support pad under oil inlet |
6. Mold Collapse (Squatting)
Mold collapse appeared as excessive flash and reduced wall thickness. We traced it to a worn scraper blade on the weight conveyor. When the weight surface had adhering sand lumps, the weight pressed unevenly into the mold. Additionally, a worn sand-drawing saw blade left the sand surface above the flask, causing the weight to crush the mold. The condition for safe pressing is:
$$ F_{\text{weight}} < A_{\text{contact}} \cdot \sigma_{\text{compressive}} $$
where \(\sigma_{\text{compressive}}\) is the compressive strength of the mold sand. Once the scraper and saw blade were replaced and included in TPM, no further collapse occurred.
| Defect | Root Cause | Countermeasure |
|---|---|---|
| Mold collapse (squatting) | Worn weight scraper; worn sand-drawing saw blade | Replace scraper and saw blade; add to TPM checklist |
7. Summary of Overall Improvement
Through these systematic corrections, our total internal and external scrap rate fell from 18% to below 2%. Table 5 consolidates all defects and the key actions.
| Defect Category | Primary Countermeasures | Before (%) | After (%) |
|---|---|---|---|
| Blowhole | Blind vents; reduced pin height; additional overflow risers | 0.45 (water pump face) | <0.015 |
| Double skin | Added blind vent pins | ~1.0 | <0.1 |
| Sand inclusion | Core handling improvements; coating control; crush rib depth reduction | ~8.0 | <0.8 |
| Cutter damage | Burr orientation change | ~2.0 | <0.1 |
| Oil inlet crack | Support pad modification | ~1.5 | 0 |
| Mold collapse | TPM on scraper and saw blade | ~0.5 | 0 |
We emphasize that a thorough understanding of the physics behind each sand casting defect is essential. For example, the venting requirement can be quantified by the following equation linking binder decomposition rate and vent cross-section:
$$ A_{\text{vent}} \geq \frac{\dot{m}_{\text{gas}}}{\rho_{\text{gas}} v_{\text{vent}}} $$
where \(\dot{m}_{\text{gas}}\) is the mass rate of gas evolution, \(\rho_{\text{gas}}\) is gas density, and \(v_{\text{vent}}\) is the allowable vent exit velocity to avoid metal penetration.
8. Lessons Learned
From our experience, we drew the following conclusions that can help any foundry dealing with gray iron cylinder blocks:
- Blind vents and overflow risers are highly effective against blowholes and double skin in thin sections.
- Complete removal of loose sand from cores and molds, uniform coating application, and clean weight surfaces are vital to prevent sand inclusions.
- All other defects can be systematically addressed by observing the production line, identifying the specific source of the problem, and applying mechanical or process changes.
- Using formulas to model gas pressure, stress, or solidification helps prioritize solutions and predict outcomes.
By sharing our journey in attacking this common sand casting defect portfolio, we hope to assist other foundry engineers in achieving similar scrap reductions. Continuous monitoring and a data-driven approach remain our key to sustaining quality.