CONTROL CHARTS FOR

VARIABLES
Dr. M. S. Memon
Dept. of Industrial Engineering and Management
Mehran University of Engineering and Technology

https://msmemon.wordpress.com/scmlab/
How to select control chart?
 Introduction

• Control charts for variables—quality characteristics that

are measurable on a numerical scale.
• Examples of variables include length, thickness, diameter,
breaking strength, temperature, acidity, viscosity, order
processing time, and waiting time for service.
• We must be able to control the mean value of a quality
characteristic as well as its variability.
• The mean gives an indication of the central tendency of a
process, and the variability provides an idea of the process
dispersion.
• Therefore, we need information about both these statistics
to keep a process in control.
 Introduction

• Let's consider Figure 7-1. A change in the process mean of

a quality characteristic (say, length of a part) is shown in
Figure 7-1 a, where the mean shifts from μ0 to μ1.
• It is, of course, important that this change be detected
because if the specification limits are as shown in Figure 7-
la, a change in the process mean would change the
proportion of parts that do not meet specifications.
• Figure 7- lb shows a change in the dispersion of the process;
the process standard deviation has changed from σ0 to σ1
with the process mean remaining stationary at μ0).
• Note that the proportion of the output that does not meet
specifications has increased. Control charts aid in detecting
such changes in process parameters.
Introduction
Introduction
 Introduction

• Variables provide more information than attributes.

• Attributes deal with qualitative information such as
whether an item is nonconforming or what the number
of nonconformities in an item is.
• Thus, attributes do not show the degree to which a
quality characteristic is nonconforming.
• For instance, if the specifications on the length of a part
are 40 ± 0.5 mm and a part has length 40.6 mm,
attribute information would indicate as nonconforming
both this part and a part of length 42 mm.
• The degree to which these two lengths deviate from the
specifications is lost in attribute information.
 Introduction

• The cost of obtaining variable data is usually higher than for

attributes because attribute data are collected by means such
as go/no-go gages, which are easier to use and therefore less
costly.
• The total cost of data collection is the sum of two
components: the fixed cost and the variable unit cost. Fixed
costs include the cost of the inspection equipment; variable
unit costs include the cost of inspecting units.
• As the use of automated devices for measuring quality
characteristic values spreads, the difference in the variable
unit cost between variables and attributes may not be much.
However, the fixed costs, such as investment costs, may
increase.
 CONTROL CHARTS FOR THE MEAN AND RANGE

• Development of the Charts

CONTROL CHARTS FOR THE MEAN AND RANGE
 CONTROL CHARTS FOR THE MEAN AND RANGE

• Rather than compute 𝜎𝑥ҧ from the raw data, we can use the
relation between the process standard deviation σ (or the
standard deviation of the individual items) and the mean of
ത
the ranges (𝑅).
• Multiplying factors used to calculate the centerline and
control limits are given in Appendix A-7.
• When sampling from a population that is normally
distributed, the distribution of the statistic W = R/σ (known
as the relative range) is dependent on the sample size n.
• The mean of W is represented by d2 and is tabulated in
Appendix A-7. Thus, an estimate of the process standard
deviation is
CONTROL CHARTS FOR THE MEAN AND RANGE
CONTROL CHARTS FOR THE MEAN AND RANGE
 CONTROL CHARTS FOR THE MEAN AND RANGE

• Step 4: Plot the values of the range on the control chart

for range, with the centerline and the control limits
drawn. Determine whether the points are in statistical
control.
• If not, investigate the special causes associated with the
out-of-control points (see the rules for this in Previous
Chapter) and take appropriate remedial action to
eliminate special causes.
• Typically, only some of the rules are used
simultaneously. The most commonly used criterion for
determining an out-of-control situation is the presence
of a point outside the control limits.
 CONTROL CHARTS FOR THE MEAN AND RANGE

• An R-chart is usually analyzed before an 𝑋- ത

chart to determine out-of-control situations.
• An R-chart reflects process variability, which
should be brought in to control first. As shown
ത
by eq. (7-7), the control limits for an 𝑋-chart
involve the process variability and hence 𝑅.ത
• Therefore, if an R-chart shows an out-of-
ത
control situation, the limits on the 𝑋-chart may
not be meaningful.
 CONTROL CHARTS FOR THE MEAN AND RANGE

• Let's consider Figure 7-3. On the R-chart, sample

12 plots above the upper control limit and so is
out of control.
ത
• The 𝑋-chart, however, does not show the process
to be out of control.
• Suppose that the special cause is identified as a
problem with a new vendor, who supplies raw
materials and components.
• The task is to eliminate the cause, perhaps by
choosing a new vendor or requiring evidence of
statistical process control at the vendor's plant.
CONTROL CHARTS FOR THE MEAN AND RANGE
CONTROL CHARTS FOR THE MEAN AND RANGE
CONTROL CHARTS FOR THE MEAN AND RANGE
 Example 1

• Consider a process by which coils are

manufactured. Samples of size 5 are randomly
selected from the process, and the resistance
values (in ohms) of the coils are measured.
• The data values are given in Table 7-2, as are
the sample mean 𝑋ത and the range R.
 Example 1

• First, the sum of the ranges is found and then

ത We have
the centerline 𝑅.
Example 1
Example 1
 Example 1

• Observe that sample 3 is above the upper control limit on

the R-chart and samples 22 and 23 are below and above the
ത
𝑋-chart control limit, respectively.
• When the special causes for these three samples were
investigated, operators found that the large value for the
range in sample 3 was due to the quality of raw materials
and components purchased from a new vendor.
• When the special causes for samples 22 and 23 were
examined, operators found that the oven temperature was
too high for sample 22 and the wrong die was used for
sample 23.
• Remedial actions were taken to rectify these situations.
 Example 1

• With samples 3, 22, and 23 deleted, the revised

centerline on the R-chart is
Example 1
 Control Limits for a Given Target or Standard

• Management sometimes wants to specify values

for the process mean and standard deviation.
• These values may represent goals or desirable
standard or target values.
• Control charts based on these target values help
determine whether the existing process is capable
of meeting the desirable standards.
• Furthermore, they also help management set
realistic goals for the existing process.
 Control Limits for a Given Target or Standard

• Let 𝑋ത0 and σ0 represent the target values of the

process mean and standard deviation,
respectively.
• The centerline and control limits based on
ത
these standard values for the 𝑋-chart are given
by
 Control Limits for a Given Target or Standard

• Let A = 3/ 𝑛. Values for A are tabulated in

Appendix A-7.
Control Limits for a Given Target or Standard
 Control Limits for a Given Target or Standard

• We must be cautious when we interpret control

charts based on target or standard values.
• Sample observations can fall outside the
control limits even though no special causes
are present in the process.
• This is because these desirable standards may
not be consistent with the process conditions.
• Thus, we could waste time and resources
looking for special causes that do not exist.
 Control Limits for a Given Target or Standard

• On an 𝑋ത -chart, plotted points can fall outside the

control limits because a target process mean is specified
as too high or too low compared to the existing process
mean.
• Usually, it is easier to meet a desirable target value for
the process mean than it is for the process variability.
• For example, adjusting the mean diameter or length of a
part can often be accomplished by simply changing
controllable process parameters.
• However, correcting for R-chart points that plot above
the upper control limit is generally much more difficult.
 Control Limits for a Given Target or Standard

• An R-chart based on target values can also

indicate excessive process variability without
special causes present in the system.
• Therefore, meeting the target value σ0 may
involve drastic changes in the process. Such an R-
chart may be implying that the existing process is
not capable of meeting the desired standard.
• This information enables management to set
realistic goals.
 Example 2

• Refer to the coil resistance data in previous

Example 1. Let's suppose that the target values
for the average resistance and standard
deviation are 21.0 and 1.0Ω, respectively. The
sample size is 5.
 Example 2

Solution
• The centerline and the control limits for the X-
chart are as follows
Example 2
 Example 2

• Figure 7-5 shows the control chart for the range

based on the standard value. Since the control
charts were revised in first Example, we plot the
22 in-control samples and exclude samples 3,22,
and 23 because we are assuming that remedial
actions have eliminated those causes.
• Now we can see how close the in-control process
comes to meeting the stipulated target values.
 Example 2

• The process seems to be out of control with respect to

the given standard. Samples 5 and 8 are above the
upper control limit, and a majority of the points lie
above the centerline. Only six of the points plot below
the centerline.
• Figure 7-5 thus reveals that the process is not capable
of meeting company guidelines.
• The target standard deviation σ0 is 1.0.
• The estimated process standard deviation from first
Example (calculated after the process was brought to
control) is
 Example 2

• This estimate exceeds the target value of 1.0.

• Management must look at common causes to
reduce the process variability if the standard is to
met.
• This may require major changes in the methods of
operation, the incoming material, or the
equipment.
• Process control will not be sufficient to achieve
the desired target.
 Example 2

ത
• The 𝑋-chart based on the standard value is shown in
Figure 7-6. Several points fall outside the control
limits—four points below and two points above.
ത
• In first Example, the revised centerline for the 𝑋-chart
was found to be 20.864. Our target centerline is now
21.0.
• Adjusting controllable process parameters could
possibly shift the average level up to 21.0.
• However, the fact that there are points outside both the
upper and lower control limits signifies that process
variability is the issue here.
Example 2
Interpretation and Inferences from the Charts

• The difficult part of analysis is determining and interpreting

the special causes and selecting remedial actions.
• Effective use of control charts requires operators who are
familiar with not only the statistical foundations of control
charts but also the process itself.
• They must thoroughly understand how the different
controllable parameters influence the dependent variable of
interest.
• The quality assurance manager or analyst should work
closely with the product design engineer and the process
designer or analyst to come up with optimal policies.
Interpretation and Inferences from the Charts

• In Chapter 4 we discussed five rules for determining

out-of-control conditions. The presence of a point
falling outside the 3σ limits is the most widely used of
those rules.
• Determinations can also be made by interpreting typical
plot patterns.
• Once the special cause is determined, this information
plus a knowledge of the plot can lead to appropriate
remedial actions.
• Often, when the R-chart is brought to control, many
ത
special causes for the 𝑋-chart are eliminated as well.
Interpretation and Inferences from the Charts

• Once a process is in statistical control, its capability can be

estimated by calculating the process standard deviation.
• This measure can then be used to determine how the process
performs with respect to some stated specification limits.
• The proportion of nonconforming items can be estimated.
• Depending on the characteristic being considered, some of
the output may be reworked, while some may become
scrap.
• Given the unit cost of rework and scrap, an estimate of the
total cost of rework and scrap can be obtained.
Interpretation and Inferences from the Charts

• From an R-chart that exhibits control, the process

standard deviation can be estimated as

• where 𝑅ത is the centerline and d2 is a factor

tabulated in Appendix A-7.
• If the distribution of the quality characteristic can
be assumed to be normal, then given some
specification limits, the standard normal table can
be used to determine the proportion of output that
is nonconforming.
 Example 3

Refer to the coil resistance data in Example 1.

Suppose that the specifications are 21 ± 3Ω.
a) Determine the proportion of the output that is
nonconforming, assuming that coil resistance is
normally distributed.
b) If the daily production rate is 10,000 coils and if
coils with a resistance less than the LSL cannot
be used for the desired purpose, what is the loss
to the manufacturer if the unit cost of scrap is 50
cents?
 Example 3

Solution (a)
• From the revised R-chart, we found the
centerline to be 𝑅ത = 3.50. The estimated
process standard deviation is
Example 3
 Example 3

• From Appendix A-3

we find that the
proportion of the
product below the
LSL is 0.0287, and
the proportion above
the USL is 0.0188.
• Thus, the total
proportion of
nonconforming
output is 0.0475.
 Example 3

• The daily cost of scrap is

 Home Work 2
• A pharmaceutical company is interested in controlling its filling operation.
Random samples of size 4 are selected and the fill weight is recorded. Table
shows the data for 24 samples. The specifications on fill weight are 350 ± 5
grams (g). Daily production rate is 20,000 bottles.
a) Find the trial control limits for the X-bar and R-charts.
b) Assuming special causes for out-of-control points, find the revised
control limits.
c) How many bottles are nonconforming daily?
d) If the cost of rectifying an underfilled bottle is Rs. 0.08 and the
lost revenue of an overfilled bottle is Rs. 0.03, what is monthly
revenue lost on average?
e) If the process average shifts to 342 g, what is the probability of
detecting it on the next sample drawn after the shift?
f) What proportion of the output is nonconforming at the level of
process average indicated in part (e)?
Note: Use Minitab to develop Control Charts
 Home Work 2
Table
 Example 4

• A hard-bake process is used in conjunction with

photolithography in semiconductor
manufacturing.
• We wish to establish statistical control of the flow
width of the resist in this process using 𝑥ҧ and R
charts. Twenty-five samples, each of size five
wafers, have been taken when we think the
process is in control. The interval of time between
samples or subgroups is one hour. The flow width
measurement data (in microns) from these
samples are shown in following Table.
Example 4
Example 4
 Example 4

The specification limits on flow width are 1.50 ± 0.50 microns. The control chart data
may be used to describe the capability of the process to produce wafers relative to these
specifications. Assuming that flow width is a normally distributed random variable,
with mean 1.5056 and standard deviation 0.1398, we may estimate the fraction of
nonconforming wafers produced as
 Example 4

That is, about 0.035 percent [350 parts per million (ppm)] of the wafers produced
will be outside of the specifications.

Another way to express process capability is in terms of the process capability ratio
(PCR) Cp (See Process Capability Slides for details)
Example 4
Example 4
Example 4
 CONTROL CHARTS FOR INDIVIDUAL UNITS

MR- Chart
• For some situations in which the rate of production is
low, it is not feasible for a sample size to be greater
than 1.
• Additionally, if the testing process is destructive and the
cost of the item is expensive, the sample size might be
chosen to be 1.
• Furthermore, if every manufactured unit from a process
is inspected, the sample size is essentially 1.
• Service applications in marketing and accounting often
have a sample size of 1.
 CONTROL CHARTS FOR INDIVIDUAL UNITS

• In a control chart for individual units—for

which the value of the quality characteristic is
represented by X—the variability of the
process is estimated from the moving range
(MR), found from two successive
observations.
• The moving range of two observations is
simply the result of subtracting the lesser
value.
MR- Chart
MR- Chart
MR- Chart
 Example 5

• Following Table shows the Brinell hardness numbers of 20 individual steel

fasteners and the moving ranges. The testing process dents the parts so that they
cannot be used for their intended purpose. Construct the X-chart and MR-chart
based on two successive observations. Specification limits are 32 ± 7.
 Example 5

Solution
Note that there are 19 moving-range values for 20 observations. The average of
the moving ranges is
Example 5
 Example 5

• The centerline of the X-chart is

 Other Control Charts

• Standard deviation chart (s-chart) - preferable for larger sample

sizes (equal to or greater than 10, usually).
• Moving range (MR-chart)
• Standardized control chart for individuals (Z-chart) – for short
production run
• cumulative sum chart
• Exponentially Weighted Moving-Average or Geometric Moving-
Average Control Chart (EWMA- chart)
• Trend Chart (Regression Control Chart)
• Modified Control Chart
• Acceptance Control Chart
• Multivariate Control Charts
• Hotelling's T2 Control Chart
• Generalized Variance Chart