The discussions in the previous sections outlined a system for investigating and maintaining process capability for the purpose of quality planning. In the six sigma environment, process capability data will have to be maintained within one or more of the indicators that were discussed in previous chapters, including DPU (PPM), DPMO, yield, and number of sigma’s quality (including six sigma). Knowing that all of these indicators are related to each other as discussed and shown by examples in previous chapters, an enterprise can decide on one of these indicators, or a combination of several, and use the indicators) in process capability studies. This is especially useful when the enterprise management or major customers have asked for a certain level of quality.
An example would be a factory that chose Cpk as the process capability indicator. This requires that all of the fabrication and assembly operations, as well as major part suppliers and outside manufacturing contractors, are to report on their process capabilities. For the suppliers and contractors, a supplier management team and contractual processes with quality as well as cost and delivery requirements have to be in place to indicate the need for process capability. The purpose of these activities is to inform the new product design teams of the current quality status of different operations in manufacturing and the supply chain. If the design team finds the process capability inadequate, manufacturing has to purchase better-quality equipment or select new suppliers that can meet the quality goals. The process capability data has to be updated regularly in order to keep design team abreast of quality and capability enhancements. The frequency of updates should be short enough to comfortably fit inside the new product design cycles, as well as meet yearly management goals. A typical frequency of updating process capability is every quarter.
For assembled parts, the process capability determination has to be compatible with industry standards, as well as the calculations of defect opportunities. For PCBs and their terminations, standards such as DPMO are used (see Section 4.3.3). For fabricated parts, especially those made in machine shops, the process capability determination is more difficult. The machine shop can produce parts with the desired geometry using many possible machines in the shop; some producing high-quality parts and others parts of much lower quality. The dilemma is whether a particular process should be machine dependent, especially since the machine selection is usually not included in the part or assembly documentation. If a V2 hole needs to be drilled, there are many alternative machines in the shop to perform this operation, with varying process capabilities. So what will the design team assume for the holes defect rate?
One solution to the fabrication dilemma is to allow for an additional attribute in the six sigma methodology. This attribute would be a quality or complexity indicator. The fabrication shop could be divided into several (maximum of three) levels of complexity. As each new part is being designed, the design engineer can select from any of the three process capabilities available, depending on the level of complexity of the part.
For each process, a baseline process capability is determined, according to the sampling methods outlined in Table 5‘4. Every quarter, all of the process capabilities are checked, and recalculated if they show a statistically significant shift in average or CT using* statistical comparison tests. The 2 distribution is used to compare a large (>30) sample with the baseline population averages; the x2 test is used to compare sample to population <J. For smaller-size samples，the sample average shift to the population average can be tested with the t distribution，as shown earlier in this chapter.
Some of the process capability data can be obtained from control charts, as shown in Chapter 3, whereas others can be calculated directly by taking samples from the production line. Table 5.5 is an ex- ample of a production line of PCB assembly process capability calculations using Cpk. It shows the baseline and the present quarter performance. The data could also be plotted versus time, with the management goals shown prominently on the graph plots.
Table 5.5 shows a process capability, measured in Cpk, for each step of the process. The process capability is checked each quarter, and the source of the check is shown. Some checks are performed by using existing control charts, including moving range (MR) charts, whereas others are checked using sampling methods. Note that two process capabilities had to be changed, since the quality performance has changed dramatically.