Contents
Quality parameters are the measurable characteristics that decide whether a product meets its requirements, together with the process settings that produce them. A shaft diameter, a seal strength or a tablet weight are quality parameters of the product. Melt temperature, spindle speed or sealing pressure are parameters of the process that drive them. Each parameter has a target value and limits, and each needs a way to be measured.
Product parameters and process parameters
The distinction matters because they are controlled in different places. Product parameters are checked on the part, after the fact. Process parameters are set and monitored on the machine, while the part is being made.
| Process | Process parameter | Product parameter it drives |
|---|---|---|
| Injection molding | melt temperature, holding pressure, cooling time | part weight, dimensions, sink marks |
| Machining | spindle speed, feed rate, tool wear | diameter, surface roughness |
| Heat sealing in packaging | sealing temperature, pressure, dwell time | seal strength, leak tightness |
| Tablet compression | compression force, press speed | tablet weight, hardness |
If the link between the two is known, a drifting process parameter warns before the product goes out of tolerance. Process monitoring is built on that link. For the measures used at the product end, see quality metrics; for machine settings in general, see production parameters.
Specification limits and control limits
Two kinds of limits are often mixed up. Specification limits come from the customer or the drawing and say what the product must meet. Control limits come from the process itself and say how much it varies when nothing unusual happens. The NIST/SEMATECH e-Handbook defines process capability as a comparison between the output of an in-control process and the specification limits (NIST).
| Specification limits | Control limits | |
|---|---|---|
| Set by | customer, drawing, standard | the process data |
| Answer the question | Is this part acceptable? | Is the process behaving as usual? |
| Used in | inspection, release | a control chart, process monitoring |
| A point outside means | a nonconforming part | a special cause to investigate, even if the part is still in spec |
Worked example: Cp and Cpk of a shaft diameter
Cp relates the width of the specification to the spread of the process, six standard deviations, and ignores where the process is centered. Cpk takes the centering into account: it divides the distance from the mean to the nearer specification limit by three standard deviations. Cpk can never be larger than Cp (NIST).
The numbers are invented for illustration. A shaft is specified at 20.00 mm ± 0.05 mm, so the limits are 19.95 and 20.05 mm. Measured parts have a mean of 20.01 mm and a standard deviation of 0.012 mm.
| Value | Calculation | Result |
|---|---|---|
| Cp | (20.05 − 19.95) ÷ (6 × 0.012) | 1.39 |
| Distance to upper limit | 20.05 − 20.01 | 0.04 mm |
| Distance to lower limit | 20.01 − 19.95 | 0.06 mm |
| Cpk | 0.04 ÷ (3 × 0.012) | 1.11 |
The spread alone would allow a Cp of 1.39. Because the mean sits 0.01 mm above the target, the process is closer to the upper limit and Cpk drops to 1.11. Centering the process at 20.00 mm would bring Cpk up to Cp without reducing variation at all. To run the numbers for your own process, use the Cpk calculator. Which minimum value applies is set by the customer or the industry. NIST uses an example in which a Cp of 1.0 with an off-center mean gives a Cpk of 0.67, and recommends reaching a Cpk of at least 1.0 by reducing spread or correcting the center.
Common mistakes with quality parameters
- Too many parameters. Measuring everything dilutes attention. Focus on the parameters that actually drive customer-relevant characteristics.
- Limits copied from the drawing to the machine. A process parameter needs its own window, derived from its effect on the product.
- Capability from too few parts. A Cpk from twenty consecutive parts covers a short run under one set of conditions. NIST considers about 50 independent values large enough and calls for at least 100 in a capability study. The data should span shifts, material lots and tool changes.
- Capability of an unstable process. Cp and Cpk assume an in-control process. If the control chart shows special causes, fix those first; see statistical process control.
- Measurement system ignored. If the gauge varies as much as the process, the numbers describe the gauge.
Quality parameters in SYMESTIC
SYMESTIC records quality at the workstation and on the process side. Operators scan the part barcode, rate each part OK or NOK with its serial number and pick the defect reason from a catalog. Reports show the top defect reasons, the defect rate in percent or PPM and its trend, and first pass yield for serialized parts. For a numeric process parameter, the process stability widget shows control limits and Cp and Cpk per machine (a segment in the system) and period.
See the page on quality control software for details, or read more on process quality.
Sources
- NIST/SEMATECH e-Handbook of Statistical Methods: 6.1.6 What is Process Capability?
A free live demo based on your own use case shows how quality inspections and process stability are tracked per part and per process parameter.
Track defect rates and Cpk per process parameter
In a free 30-minute live demo, see how parts are rated OK or NOK at the workstation and how Cp and Cpk are shown per process parameter.
- Live OEE
- Downtime and reasons
- Connecting machines and ERP