x bar, R and P charts
Dr A R M Harunur Rashid
History
• This general theory of control charts was first proposed by
Walter A. Shewhart in Bell Laboratories, USA, and control
charts developed according to these principles are often
called Shewhart control charts.
In the general model for a control chart. Let w be a sample
statistic that measures some quality characteristic of
interest, and suppose that the mean of w is muw and the
standard deviation of w is sigmaw. Then the center line, the
upper control limit, and the lower control limit make the
control chart where L is the “distance” of the control limits
from the center line, expressed in standard deviation units.
68–95–99.7 rule
The very data used to produce chart
First must be tested?
• Yes, to make the chart reliable. Say you
take samples in last month and produce a
control chart. Now first thing you have to
do check that last month’s sample data
things are within control in this chart. If not,
you could not use this chart for subsequent
use. You need to stop production and find
out reasons like 3M- man, machine and
material.
Why process control?
• Find out natural quality level of a group of
staff with the same machine or just
individual.
• Then in the subsequent days, if the group
goes out of control, that it means not
natural. So reasons-man,machine and
material to be found out as such occurrence
of out of control is unnatuaral.
Why process control?
• The process control chart will be posted on the shop floor.
So the group will try to improve its quality. Look
Besterfield.
• For example, UCL of defective percent may go down from
this month(March) to next month(April), i.e. say
CL=1.27% UCL=3.25%. In the next month when we
consider the data of previous month as samples, we found
CL=1%,UCL=2.5%. So the group has improved their
natural quality level by increased effort or training or
improved machines or improved process or so on.
• What gets measured get done, or what get
measured get improved.
• So we have the evaluation process like
exams.
• TP Bolts Ltd. Produces bolts of a certain diameter. From
a day’s production a sample of 5 bolts is selected
randomly from the production line and their diameters
(in mm) are recorded. The average diameter and range
of this sample (of size 5) are computed and recorded.
The Quality Control Engineer collected this type of
samples in 10 days in the month of February and the
findings are shown in table below. From this table, draw
the X-bar and R chart. After finalizing the control
charts, in a given day, five(5) bolts are randomly
selected with the diameter(mm) of 10.746, 10.789,
10.01, 9.89, and 10.20. Now using the control charts,
comment on it.
Day Average diameter of Range, R
the sample(mm)
1 10.769 0.050
2 10.730 0.016
3 10.718 0.040
4 10.728 0.014
5 10.730 0.029
6 10.720 0.020
7 10.711 0.038
8 10.713 0.026
9 10.718 0.008
10 10.789 0.032
P chart formula
Q2. Draw the P chart using the above data. Now if you have taken a
sample of 200 cards today and found the percentage of cards in error as
4 percent. So what would you do as a production engineer?
• Percentages are to be converted to decimal
values and these equivalent decimal values
are to be used in the P chart equations.
• Quantitative : x bar and R chart
• Qualitative: p chart
• Conservative mode: if several control charts
are used in the process and any one of the
charts signals out of control, production
may be stopped and an investigation may be
carried out.