In the landscape of modern, large-scale rubber industry production, particularly for critical processes like manufacturing all-steel radial tire compounds, the twin-screw extruder stands as a pivotal piece of equipment. Its performance and reliability are paramount. At the heart of this machine lies a key component: the screw. Subjected to extreme mechanical stresses and abrasive wear, the screw demands exceptional strength and durability. To meet these requirements, casting is the preferred manufacturing method. However, this process is notoriously prone to introducing various casting defects, which historically led to low pass rates, costly rework, delayed deliveries, and significant frustration for machining teams who bore the brunt of processing flawed castings. This article details a first-person account of a systematic project undertaken to radically improve the quality of these critical castings by employing Lean Six Sigma methodologies to understand, analyze, and control the root causes of these persistent casting defects.
The initial situation was untenable. A significant portion of screw castings, after initial rough machining, required return to the foundry for extensive repair welding—a process demanding specialized procedures and validation. This rework loop increased workload, multiplied handling, and critically jeopardized production schedules. We defined our problem clearly using a Lean metric: the Casting Defect Rate.
$$
\text{Casting Defect Rate} = \frac{\text{Number of Defective Screw Castings}}{\text{Total Number of Screw Castings}} \times 100\%
$$
Analysis of historical data from a twelve-month period established a distressing baseline. The average casting defects rate was calculated at 63.8%, with monthly figures fluctuating dramatically, as shown in the consolidated data below. Our goal was to reduce this rate to a target of 34%.
| Period | Casting Batches Analyzed | Baseline Avg. Defect Rate | Project Target Defect Rate |
|---|---|---|---|
| 12 Months (Historical) | 61 | 63.8% | 34.0% |
A deeper dive into the nature of these casting defects was essential. We categorized and quantified the flaws discovered during machining over a six-month period. The data revealed a clear hierarchy of problems.
| Defect Type | Frequency | Percentage of Total Defects |
|---|---|---|
| Shrinkage Porosity | 28 | 39.4% |
| Gas Porosity (Blowholes/Pinholes) | 27 | 38.0% |
| Hot Tears/Cracks | 12 | 16.9% |
| Sand Inclusions | 2 | 2.8% |
| Miscellaneous (Miss-runs, etc.) | 2 | 2.8% |
The predominance of shrinkage, gas porosity, and cracks—accounting for over 94% of all casting defects—provided a focused direction for our investigation. We mapped the entire manufacturing process and conducted a cause-and-effect analysis (Ishikawa diagram) to brainstorm potential sources of variation contributing to these flaws. The major categories examined included Methods, Materials, Machinery, Manpower, and the Environment within the foundry process flow: Pattern Making, Molding, Melting, Pouring, and Heat Treatment.
To prioritize our efforts, we performed a Failure Mode and Effects Analysis (FMEA). This rigorous method assessed each potential failure mode for its Severity (SEV), Occurrence (OCC), and Detectability (DET), calculating a Risk Priority Number (RPN = SEV × OCC × DET). High RPN items became our primary targets. The FMEA table was instrumental in transitioning from generic causes to specific, actionable process inputs.
| Process Step | Key Input | Potential Failure Mode | Potential Defect | SEV | OCC | DET | RPN |
|---|---|---|---|---|---|---|---|
| Pattern Design | Riser Size & Design | Insufficient Modulus for Feeding | Shrinkage Porosity | 10 | 8 | 5 | 400 |
| Molding | Internal Chill Cleanliness | Surface Contamination (Rust, Scale) | Gas Porosity, Inclusions | 9 | 7 | 6 | 378 |
| Mold Drying | Mold Drying Time/Temp | Insufficient Drying | Gas Porosity | 9 | 7 | 6 | 378 |
| Melting | Chemical Composition | Excessive S, P Content | Cracks (Hot Tearing) | 10 | 3 | 4 | 120 |
The analysis phase yielded immediate “Quick Win” opportunities—factors we could address rapidly with existing resources and strict procedural controls. Implementing these controls formed the first wave of our countermeasures against casting defects.
| Area | Key Factor | Previous State | Improvement Action |
|---|---|---|---|
| Materials | Sand Properties (Moisture, Clay, Grain Fineness) | Spot checks only; no strict inbound control. | Implemented 100% inspection per delivery batch. Records maintained. Non-conforming sand rejected. |
| Alloying/Deoxidizing Additives | Materials sometimes used damp from storage. | Mandated pre-heating for all additives to remove moisture before charging into the furnace. | |
| Molding Process | Riser & Gating Placement/Size | Relied on operator skill; minimal in-process verification. | Strict adherence to documented placement. Use of high-insulation sleeves for risers. Enhanced supervisor checks and audits. |
| Internal Chill Preparation | Manual cleaning inconsistent. | Implemented standardized shot blasting of chill assemblies to achieve a clean, “white metal” surface prior to placement in the mold. | |
| Melting & Pouring | Mold Drying | Inconsistent time/temperature control. | Optimized cycle to 6 hours at 250°C with automated furnace recording. |
| Deoxidation Practice | Aluminum addition at 0.8%. | Increased Aluminum deoxidizer addition to 1.0% to improve metal cleanliness and reduce gas-related casting defects. | |
| Pouring Temperature & Practice | Infrequent temperature measurement; inconsistent riser feeding. | Mandatory pyrometer check per ladle. Defined minimum holding time. Enforced “topping up” of risers after pouring. | |
| Process Control | Cooling & Heat Treatment | Unrecorded pour-to-shakeout time; unverifiable annealing cycles. | Mold marking with pour time to enforce minimum cooling. Repair of annealing furnace recorders to validate time-temperature profiles. |
The effect of these initial countermeasures was significant, particularly on gas porosity. A comparative study showed a drastic reduction. However, shrinkage porosity remained a stubborn challenge, indicating that our initial quick wins, while necessary, were not sufficient to address the fundamental design limitation. The root cause traced back to the pattern itself. The original machining allowance was excessive (23mm), leading to an unnecessarily heavy casting section that demanded a riser of impractical size to effectively feed and compensate for solidification shrinkage—a direct recipe for shrinkage cavity casting defects.
This called for a second, more fundamental wave of improvement. We launched a cross-functional team with foundry and machining engineers. By analyzing machined castings, we determined a 13mm allowance was sufficient. Redesigning the pattern with this reduced allowance decreased the casting weight by approximately 250kg. This had a profound effect: the solidification modulus of the casting was reduced, making it easier for a properly sized riser to feed the section effectively. The relationship between riser efficiency, casting modulus, and shrinkage can be conceptually framed by Chvorinov’s Rule, which states that solidification time is proportional to the square of the volume-to-surface area ratio (the modulus):
$$
t_s = k \left( \frac{V}{A} \right)^2
$$
Where \( t_s \) is solidification time, \( V \) is volume, \( A \) is surface area, and \( k \) is a mold constant. By reducing the volume \( V \) of the casting section, its solidification time decreased relative to the riser, improving feeding efficiency and directly combating the casting defects of shrinkage porosity. The new pattern, combined with the established process controls, was put into production.
The results from the subsequent production batches were transformative. The data below compares defect instances before and after the full implementation of both procedural controls and the pattern redesign.
| Metric | Before Improvements (Baseline) | After Quick Wins | After Full Implementation (Pattern + Controls) |
|---|---|---|---|
| Casting Defects Rate (Target: 34%) | 63.8% | ~50% (Est. post-initial fixes) | 33.4% |
| Gas Porosity Defect Rate | ~38% (of total defects) | 16.7% (of castings) | 16.7% (of castings, & vastly reduced severity) |
| Shrinkage Porosity Defect Rate | ~39% (of total defects) | High (No significant change) | 16.7% (of castings, & reduced depth/severity) |
| Cracking Defect Rate | ~17% (of total defects) | Reduced | 0% |
| Scrap Rate | Present | Present | 0% |
The final integrated solution successfully met and slightly exceeded the 34% defect rate target. More importantly, the nature of the remaining casting defects shifted from severe, repair-intensive flaws to minor, more manageable imperfections. The complete elimination of cracking—often a catastrophic defect leading to scrap—was a major victory, directly linked to the reduction in severe shrinkage that acted as a stress raiser.
In conclusion, this journey underscored the power of a structured, data-driven approach to solving complex manufacturing problems like pervasive casting defects. By applying Lean Six Sigma tools—from precise metric definition and process mapping to root cause analysis via FMEA—we moved from reactive firefighting to proactive process control. The solution was not a single silver bullet but a combination of enhanced procedural discipline and intelligent design modification. The dramatic improvement in screw casting quality delivered cascading benefits: reliable machining schedules, predictable costs, on-time delivery of extruders, and ultimately, higher customer satisfaction. This case also serves as a valuable template for tackling quality challenges in other complex casting applications, demonstrating that rigorous methodology can yield exceptional results in even the most demanding foundry environments.