Steps on How to Set Up a Capability Study
Before we set up a capability study, we must select the critical dimension or variable to be examined (see Chapter 2 for more information). This dimension must meet product specifications. In the simplest case, the study dimension is the result of a single, direct process. In more complicated studies, the critical dimension may be the result of several processes. It may become necessary in these cases to perform capability studies on each process. Studies of early processes prove to be more valuable than studies of later processes since early processes lay the foundation that may affect later operations.
Once the critical dimension is selected, data measurements can be collected. This can be accomplished manually or by using automatic gaging and fixturing linked to a data collection device. When collecting measurements on a critical dimension, it is important that the measuring instrument be as precise as possible, preferably one order of magnitude finer than the specification. Otherwise, the measuring process will contribute excess variation to the dimension. Using handheld data collectors with automatic gages will help reduce errors introduced by the measurement process, data recording, and transcription for computer post-processing.
The ideal situation for data collection is to collect as much data as possible over a defined time period. This will yield a capability study that is very reliable, since it is based upon a large sample size.
In the 6 steps of process improvement, determining process capability is Step 5:
1) Gather process data (Steps 1 through 4 are covered in Chapters 1 through 3).
2) Plot the data on control charts.
3) Find the control limits.
4) Identify and eliminate assignable causes.
5) Calculate process capability.
6) If process capability is not sufficient, reduce the random cause variation,
and go back to Step 1.
The Process Must be in Control
The process must be in control and normally distributed before samples can be taken to measure process capability. All standard capability indices assume that the process is in control and the individuals follow a normal distribution. If the process is not in control, capability indices are invalid, even if they indicate the process is capable.
There are three statistical tools used to determine whether the process is in control and follows a normal distribution (No single tool can be used; they must all be used together):
- Control charts
- Visual analysis of a histogram
- Mathematical analysis of the distribution
Control charts are the most common method for operators to maintain statistical control of the process. With control charts, such as an x̄ & R chart, all points must be inside the control limits with no apparent patterns or trends present (see Chapter 3 for details).
Chapter 2 explains how to create an histogram. An histogram allows us to see if any parts are outside the specification limits and what the distribution’s position is relative to the specifications. If the process is naturally a normal distribution, then the histogram should approximate a bell-shaped curve when the process is in control. A process can be in control without having its individuals following a normal distribution if the process is inherently abnormal.
See 5.5 Tests for Normality for statistical calculations that assist in identifying whether the process is normal
The Process Must be Inherently Normal
Many processes naturally follow a bell-shaped curve (a normal distribution), but some do not. Examples of abnormal dimensions are roundness, squareness, flatness, and positional tolerances; they have a natural barrier at 0. In these cases, a perfect measurement is 0 (for example, no ovality in the roundness measurement). There can never be a value less than 0. The standard capability indices are not valid for such abnormal distributions. See 5.5 Tests for Normality and 5.6 Capability Measures for Abnormal Distributions for more information.