Chapter 10 – Quality Control
The purpose of quality control is to ensure that processes are performing in an acceptable manner.
Quality Control: An activity that evaluates quality characteristics relative to a standard, and takes corrective
action when they do not meet standards.
Accomplished by monitoring and inspecting the product during process
The best companies emphasize designed quality into the process, thereby reducing the need for inspection/tests
or control efforts. The most progressive companies have achieved an inherent level of quality that is sufficiently
high that they can avoid wholesale inspection/test and process control. That is the ultimate goal.
Acceptance sampling – quality assurance that relies primarily on inspection/tests of previously produced items
Inspection: appraisal/comparison of goods or services against standards
The purpose of inspection is to provide information on the degree to which items conform to a standard. There
are various types of testing equipment and software available.
Some Basic decision making issues are;
How much to Inspect
The amount of inspection can range from no inspection, to inspection of each item.
Low cost high volume items require little inspection, items that have a large costs associated with passing
defective products require inspection. In high volume systems, automated inspection takes place
Amount of inspection needed is governed by costs of inspection & expected costs of passing defective items.
Traditional View: the amount of inspection is optimal when the sum of the costs of inspection and passing defectives is minimized.
As a rule, operations with a high proportion of human involvement necessitate more inspection effort than
Where to Inspect in the Process
Raw materials and purchased parts
Before a costly operation
Before an irreversible process
Before a covering process
Centralized vs. On-Site Inspection
Some situation require that inspection be performed on site. ( inspecting the hull of a ship for cracks requires
inspectors to visit the ship)
At other times, specialized tests can be performed in a lab (medical tests, analyzing food samples)
The central issue in the decision concerning on-site or lab inspections is whether the advantages of specialized
lab tests are worth the time and interruption needed to obtain the results.
STATISTICAL PROCESS CONTROL
Statistical Process Control: Statistical evaluation of the output of a process during production.
They take periodic samples from the process and compare them with predetermined limits
The Quality Control Steps
- Define the quality characteristics in detail (e.g. color, texture) to monitor
- Measure the characteristics (only the characteristics that can be measured are candidate for control)
- Compare to a standard and evaluate (the level of quality being sought)
- Take corrective action if necessary
- Evaluate corrective action
Types of Variations
Random/Common variation : Natural variations in the output of process, created by countless minor factors
Assignable/Special variation: A variation whose source can be identified
The main task in quality control is to distinguish assignable from random variation. This is done by taking a
sample of two or more observations, calculating a sample statistics such as sample mean, and using this sample
statistic makes the task easier
Control Charts: A time ordered-plot of a sample statistic with limits, used to distinguish between random and
non-random variability. Control Limits: The dividing lines between random and nonrandom deviations from the mean of the sampling
Control charts have two limits that separate random variation and nonrandom variation. The larger value is the
upper control limit (UCL) and the smaller value is the lower control limit (LCL). A sample statistic that falls in
between these two limits suggests randomness, while a value outside of, or on either one of these limits suggests
Type I error: concluding that a process has changed when it has not
Type II error: concluding a process is in control when it is actually not
Each sample mean is compared to the extremes of the sampling distribution (the control limits) to judge if it is
within the acceptable range.
Designing Control Charts
1. Determine a sample size (usually between 2- and 20, the larger the n, the smaller probability of Error)
2. Obtain 20 to 25 samples
3. Establish preliminary control limits and graph them using the formulas, and then graph them
4. Plot the sample statistic values on a control chart, and note whether any points fall outside control limits
5. If no points outside CL assume there is no assignable cause. If points are outside CL investigate &
correct 6. Operators use control chart by recording the value of sample statistic which is periodically taken
Sample Mean Control Chart: used to monitor the mean of a process (x bar chart)
If a reason estimate of the standard deviation is available, one can calculate control limits using these formulas;
UCL = + zσ x σx is .σ . z is the standard normal deviate (usually 3)
LCL = - σ x
Another approach to calculating control limits is to use the sample range (max value – min value in sample) as a
measure of process variability
UCL = + A2 A2is from table 10-2 using n
LCL = - A2 is the average of sample ranges
Sample Range Control Chart: used to monitor process dispersion
UCL = D4 where the values oD 4and D 3are from table 10-2 using
LCL = D
Individual Unit and Moving Range Charts
When the rate of production is low, testing is expensive, or there is no reason to expect additional information
by taking more observations, only one unit is sued for inspection. In this case, the sameple mean control chart
reduces to X-chart.
UCL = + zσ LCL = - zσ
However, the “sample” range cannot be calculated in the usual way because there is only one observation in
each sample. Instead, the differences between consecutive observations, (