標籤: process capability

Holding a tolerance on the first few parts is easy; holding it across a hundred thousand is a different problem entirely. Statistical process control, or SPC, is the method that keeps a molding or stamping process centered and predictable over a long run, and process capability indices such as Cpk are how that predictability is expressed as a single number a buyer can act on. For OEM purchasers who need parts that assemble reliably batch after batch, an experienced Taiwan mold maker that practices genuine SPC provides more than a passing inspection report: it provides evidence that the process itself is stable and capable of staying within specification.

INTERTECH has more than 30 years of experience in tooling and production, all 100% made in Taiwan. This article explains how SPC works, what control charts reveal, the difference between Cp and Cpk, what capability targets mean in practice, and how disciplined process control across molding and stamping gives buyers confidence in every shipment.

Why Process Control Matters More Than a Final Sort

Inspecting parts at the end and throwing away the bad ones is expensive and, on its own, unreliable. Every scrapped part represents wasted material, machine time, and energy, and a final sort based on sampling can still let defects slip through. The better approach is to build quality into the process so that bad parts are rarely made in the first place. SPC does exactly this by monitoring the process while it runs, detecting shifts and trends early, and prompting correction before parts drift out of tolerance. The goal is prevention rather than detection, and for high-volume molded and stamped parts that shift in the economics is decisive.

This matters especially because molding and stamping processes naturally drift. Melt temperature fluctuates, material lots vary, tools warm up and wear, and press conditions shift over a shift. SPC does not eliminate this variation; it makes the variation visible and separates the normal, expected noise of a stable process from the special causes that signal something has genuinely changed.

Common and Special Cause Variation

The conceptual foundation of SPC is the distinction between two kinds of variation. Common cause variation is the inherent, random scatter present in any stable process; it is the background noise you cannot economically remove without changing the process itself. Special cause variation is a signal: a new material lot, a worn tool feature, a setup error, or a machine fault that pushes the process off its usual behavior. Reacting to common cause noise as if it were a signal, sometimes called tampering, actually increases variation. SPC gives operators an objective rule for telling the two apart, so they adjust when they should and leave the process alone when they should.

How Control Charts Work

The everyday tool of SPC is the control chart, which plots a measured characteristic over time against a centerline and statistically derived control limits. Those control limits are calculated from the process’s own variation and are distinct from the specification limits on the drawing. As long as points fall randomly within the control limits, the process is behaving as expected. Certain patterns raise a flag that a special cause is present and action is needed.

• A point beyond an upper or lower control limit signals that the process has shifted outside its normal behavior.
• A run of consecutive points on one side of the centerline suggests a sustained shift in the process average.
• A steady upward or downward trend often indicates tool wear or a gradual change in conditions.
• Unusual cycling or hugging of a limit can point to a systematic influence that deserves investigation.

By responding to these signals promptly, an operator can correct a molding or stamping process before it produces a single out-of-tolerance part, which is the whole point of monitoring rather than merely inspecting.

Understanding Cp and Cpk

Process capability indices compress the relationship between a process’s variation and the tolerance it must live within into a single number. Cp compares the width of the tolerance band to the spread of the process, assuming the process is perfectly centered; it answers “could this process fit inside the tolerance if it were aimed dead center.” Cpk is the more honest and more useful index because it also accounts for how far the process average sits from the center of the tolerance. A process can have a healthy Cp yet a poor Cpk if it is running off-center, hugging one specification limit even though its spread is tight.

Because Cpk reflects both spread and centering, it is the index buyers and quality standards most often specify. A higher Cpk means more margin between the process and the nearest specification limit, and therefore a lower probability of producing a nonconforming part. Widely used capability targets ask for a comfortable cushion so that normal drift never reaches the tolerance edge; the exact target is agreed between customer and supplier based on the criticality of the characteristic.

Capability Studies at Launch and in Production

Capability is assessed at two moments. At launch, an initial study on parts from the production tool establishes whether the process is capable before approval, and these results feed directly into a PPAP submission for significant characteristics. In ongoing production, capability is monitored over time to confirm the process remains stable and centered. A characteristic identified as significant, whether because it affects fit, function, or a downstream regulatory requirement, is the natural focus for both the initial study and continued charting. INTERTECH aligns the inspection plan with the drawing so that the features that matter most are the ones under statistical watch.

Applying SPC to Molding and Stamping Specifically

The characteristics chosen for control reflect how each process behaves. In injection molding, dimensions sensitive to shrink and packing, along with weight and cosmetic attributes, are typical monitored characteristics, and for multi-cavity tools capability may be assessed with cavity-to-cavity variation in mind. In progressive-die stamping, feature dimensions influenced by springback and tool wear, along with burr height and flatness, are natural candidates for charting because they drift predictably as a die accumulates strokes. Understanding these mechanisms lets the process engineer set control limits and sampling frequencies that actually catch the way the specific process tends to move, rather than applying a generic template.

One-Stop Process Discipline in Taiwan

When design feedback, tooling, and production are divided among suppliers, the knowledge needed to control a process is fragmented too, and capability suffers at every handoff. INTERTECH provides a one-stop path from DFM feedback and mold making through molding, stamping, and inspection, with more than 30 years of experience and full made-in-Taiwan capability. Because the same organization designs the tool, runs the process, and charts the results, capability is engineered in from the start: tolerances are set realistically during design review, tools are built to hold them, and SPC confirms that the process stays centered across the run. That continuity is what turns a good Cpk from luck into a repeatable outcome.

What Buyers Should Evaluate

Before relying on a supplier’s process control, review the following checklist.

• Confirm the supplier practices real SPC with control charts, not just end-of-line sorting.
• Ask how significant characteristics are chosen and which ones are placed under statistical monitoring.
• Verify that initial capability studies support the PPAP submission for critical features.
• Understand the Cpk target agreed for your part and the reasoning behind it.
• Check that control limits and specification limits are treated as distinct, and that operators know how to respond to chart signals.
• For multi-cavity molds and long stamping runs, ask how cavity variation and tool wear are accounted for.

Conclusion

Statistical process control and process capability turn quality from a hope into a measurable, predictable property of the process. Control charts make variation visible and separate normal noise from real signals, while Cpk expresses how much margin a centered process holds against its tolerance. Together they let a supplier prevent defects rather than merely catch them, which is exactly what high-volume molded and stamped parts demand. A Taiwan mold maker that engineers capability in from design through production gives buyers stable processes and dependable shipments. If you are looking for a reliable injection mold maker in Taiwan that practices disciplined SPC and process capability for your project, please contact INTERTECH to discuss your drawings, materials, and production requirements.

Related Articles

• Quality Systems for Molding and Stamping
• Medical Molding Quality (ISO 13485)
• UL Certification for Molded Components