The Control Chart:
Shewhart control chart is an important tool in statistical quality control. Measured quality of manufactured product is always subject to a certain amount of variation as a result of chance. Some stable "system of chance causes" is inherent in any particular scheme of production & inspection. Variation within this stable pattern is inevitable. This reasons for variation out side this stable pattern may be discussed & corrected. The power of the Shewhart technique lies in its ability to separate out these assignable causes of equality variation. This makes possible the diagnosis and correction of many production troubles and often brings substantial improvements in product quality & reduction of spoilage & rework. Control charts permits better decisions on engineering tolerances and better comparisons between alternative designs and between alternative production methods. It often provides better quality assurance at lower cost.
Before production starts, a decision is necessary as to what is to be designed and made. Next is actual production and finally it must be verified whether the product manufactured is what was intended. It is convenient to think of all matters related to quality of manufactured product in terms of specifications, production, and inspection.
The techniques should be understood at a management level that encompasses all three functions. In 1980s Dr. W. Edwards Deming has been most eloquent in improving up on U S management the importance of these tools. ( of course the route was via Japan ).
There are four separate but related techniques that contributes the most common working tools in quality control.
- Control charts for measurable quality characteristics. These are described as charts for variables, or as charts for average and range and charts for average and standard deviation.
- Control chart for fraction rejected. This is described on the 'p' chart.
- Control chart for number of non-conformities per unit. This is described as 'c' chart.
- That portion of sapling theory which deals with quality protection given by any specified sampling acceptance procedure.
Control charts for variables provide information on three matters.
- Basic variability of the quality characteristics.
- Consistency of performance.
- Average level of quality characteristics.
No production process produce all items exactly alike. The merit of control chart is that it tells when to leave a process alone as well as when to correct troubles as well as when to take action to correct trouble. The elimination of assignable causes of erratic fluctuation is described as bringing process under control & is responsible for many of the cost savings resulting from statistical quality control.
Control charts for fraction rejected tell whether assignable causes of variation appear to be present or variations are explainable on chance ground. Control charts for attributes are less sensitive than the control charts for variable and does not have as great a diagnostic value. However the 'p' chart often serves to point out those situations needing diagnosis of trouble by the control chart for variables.
Control chart for non conformities per unit applies only to a limited number of manufacturing situations involving quality. It has broad applications to many other types of situations commonly met in every day life.
Acceptance Sampling:
Although statistical sampling acceptance procedures are generally superior to traditional sampling methods established without reference to the laws of probability, any one who uses acceptance sampling must face the fact that whether a portion of the stream of products submitted for acceptance does not confirm to specifications.
The greatest opportunities for cost reduction from SQC often arises out of applications of the Shewahart control chart for variables. They come from various sources - from reduction in cost of spoilage & rework, reduction in inspection cost, better control over quality of purchased product.etc.
Process Capability:
The basic statistical problem in process quality control is that of establishing a state of control over the manufacturing process i. e. eliminating special causes of variation, then maintaining that state of control through time.
Once the state of control has been established, then the question is "Is the output meeting specifications & if not, can the process be adjusted to a level where it will?
Actions that result in a change or adjustment in a process, directed at eliminating common causes, are the result of some form of Capability study.
The comparison of natural tolerance limits with specification on limits and the natural tolerance range with specification range may lead to any of the following possible course of action.
- No action is required when natural tolerance limits fall well within the specification limits.
- Action to adjust centering is required when natural tolerance range is about the same as the specification range.
- Action to reduce variability is required when cases where several products streams merge into one line prior to inspection, and in other cases analysis of source of variation may be required resulting in changes of methods, tooling, material, or equipment.
- Action to change specification: This is a design decision but one that should not be ignored by quality control personnel.
|
