SPC-Control Charts

© 2005 Wiley
Sources of Variation in Production
and Service Processes
 Common causes of variation
 Random causes that we cannot identify
 Unavoidable
 Cause slight differences in process variables like diameter,
weight, service time, temperature, etc.
 Assignable causes of variation
 Causes can be identified and eliminated
 Typical causes are poor employee training, worn tool,
machine repair, poor quality materials etc.
 Control Charts
 UCL = Process Average + 3 Standard
Deviations
 LCL = Process Average - 3 Standard
X Deviations
UCL

+ 3
Process
Average
- 3
LCL

TIME

Levine, Prentic-Hall
 Control Charts
 Graph of sample data plotted over
time
Assignable
Cause Variation
X
60
UCL
40 Process
Mean
Average 
20
LCL
0
Random
Variation 1 3 5 7 9 11
Time

Levine, Prentice-Hall
 Control Charts
X X X

Common Cause Special Cause Variation: Downward Pattern: no

Variation: no points two points outside control points outside control
outside control limit limit limit; however, eight or
more points in trend

Levine, Prentice-Hall
Process control
Measuring Variation: The
Standard Deviation
Small vs. Large
Variation
Process Capability
 A measure of the ability of a process to meet
preset design specifications:
 Determines whether the process can do what we
are asking it to do
 Design specifications (tolerances):
 Determined by design engineers to define the
acceptable range of individual product
characteristics (e.g.: physical dimensions, elapsed
time, etc.)
 Based upon customer expectations & how the
product works
Relationship between Process Variability and
Specification Width
 Types of Control Charts
 Control chart for variables are used to
monitor characteristics that can be measured,
e.g. length, weight, diameter, time, etc.
 Continuous data
 Control charts for attributes are used to
monitor characteristics that have discrete
values and can be counted, e.g. % defective,
number of flaws in a shirt, number of broken
eggs in a box, etc.
 Good or bad
Control Charts for Variables

 Mean (x-bar) charts

 Tracks the central tendency (the average
value observed) over time
 Measures the process average
 Range (R) charts:
 Tracks the spread of the distribution over
time (estimates the observed variation)
 Measures the process variation
 Data
 From Gillette example 15 hours become
15 subgroups representing the total
Data population of razor blades and each
subgroup will have 5 data values n = 5.
 Table of Data from Gillette
Example 5 measurements for each
subgroup

Subgroups
Calculations for the Data
 Find the mean (X) and the range for
each subgroup
1. X = ( Σ xi ) / n
2. R = Max (xi) – Min (xi)
 Find the mean of the X and Range
1. X = ( Σ X i ) / (#of sub groups )
2. R = ( Σ Ri ) / (# of sub groups )
Data with Calculated
figures
Create Control Charts
Limits to our Control
charts
 Upper and lower control limits represent
3 standard deviations above and 3
standard deviations below the mean
line respectfully.
 Instead of calculating the standard
deviation for all of the sample data we
use these simple equations
Control Limit Equations
 X (UCLx) = X + A2*R
 X (LCLx) = X - A2*R
Where A2 can be found on page 237 of your Intro to IE
textbook
 R (UCLR) = D4*R
 R (LCLR) = D3*R
Where D4 and D3 can be found on page 237 of your
Intro to IE textbook.
 An Alternative Method for the X-bar
Chart Using R-bar and the A2 Factor
Factor for x-Chart Factors for R-Chart
Sample Size
(n) A2 D3 D4
Use this method when 2 1.88 0.00 3.27
sigma for the process 3 1.02 0.00 2.57
distribution is not 4
5
0.73
0.58
0.00
0.00
2.28
2.11
known. Use factor A2 6 0.48 0.00 2.00
from Table 6.1 7 0.42 0.08 1.92
8 0.37 0.14 1.86
9 0.34 0.18 1.82
10 0.31 0.22 1.78
11 0.29 0.26 1.74
12 0.27 0.28 1.72
13 0.25 0.31 1.69
14 0.24 0.33 1.67
15 0.22 0.35 1.65
Finished Control Charts
Control Chart Analysis
 The data is represented over six standard
deviations,
 Three standard deviations from the mean line
to the upper control limit and three from the
mean to the lower control limit.
 The charts is divided into six sections A, B,
and C representing the standard deviations.
Example of Different
Zones
Process out of Control
1. If one point falls outside of the 3
sigma control limits (beyond zone A)
Process out of Control
2. If two out of any three successive
points fall in zone A of the same side
Process out of Control
3. If four out of any five successive
points fall in or beyond zone B of the
same side
Process out of Control
4. If eight successive points fall in or
beyond zone C of the same side
How does the Co-ops Control
Charts Look?
Control Charts for Attributes –
P-Charts & C-Charts
 Use P-Charts for quality characteristics that
are discrete and involve yes/no or good/bad
decisions
 Percent of leaking tubes in a box of 48
 Percent of broken eggs in a carton
 Units of products failing a stress test
 Use Binomial distribution

 Use C-Charts for discrete defects when there

can be more than one defect per unit
 Number of flaws or stains in a carpet sample cut from a
production run
 Number of complaints per customer at a hotel
 Use Poisson distribution
Constructing a P-Chart:
A Production manager for a tire company has inspected the
number of defective tires in five random samples with 20
tires in each sample. The table below shows the number of
defective tires in each sample of 20 tires.

Sample Sample Number

Size (n) Defective
1 20 3
2 20 2
3 20 1
4 20 2
5 20 1
Step 1:
Calculate the Percent defective of Each Sample
and the Overall Percent Defective (P-Bar)

Sample Number Sample Percent

Defective Size Defective
1 3 20 .15
2 2 20 .10
3 1 20 .05
4 2 20 .10
5 1 20 .05
Total 9 100 .09
Step 2: Calculate the Standard
Deviation of P.

p(1-p) (.09)(.91)
σp = = =0.064
n 20
 Step 3: Calculate CL, UCL, LCL

 Center line (p bar):

CL  p  .09
 Control limits for ±3σ limits:

UCL  p  z σ p   .09  3(.064)  .282

LCL  p  z σ p   .09  3(.064)  .102  0
Step 4: Draw the Chart
NP chart
 Its same as like P chart.
 It uses number of nonconforming items
rather than fraction nonconforming

UCL  np  3 np(1  p)
CL  np
LCL  np  3 np(1  p)
C-Chart
 In a situation an item may not conform
several specification but may still perform
adequately
 Few minor non conformities may be tolerated
 Number of air bubbles in a glass bottle
 Number of defective welds per meter of pipe
line
Constructing a C-Chart:

Week Number of
The number of Complaints
weekly customer 1 3
complaints are 2 2
monitored in a 3 3
large hotel. 4 1
Develop a three 5 3
sigma control limits 6 3
For a C-Chart using 7 2
the data table On 8 1
the right. 9 3
10 1
Total 22
 Calculate CL, UCL, LCL

 Center line (c bar):

#complaints 22
CL    2.2
# of samples 10

 Control limits for ±3σ limits:

UCL  c  z c  2.2  3 2.2  6.65

LCL  c  z c  2.2  3 2.2  2.25  0
U-chart
 Same as that of C-Chart (sample size is
same)
 Its if used for varied sample size
u
UCL  u  3
n
CL  u
u
LCL  u  3
n
 Control Chart Decision Tree
Defective Units Is the size of No, varies Use p-Chart
the inspection
(possibly with multiple defects) sample fixed?
Binomial Distribution
Yes, Use np-Chart
constant

Discrete What is the Individual Defects Is the size of Yes, Use c-Chart
inspection the inspection constant
Attribute basis? Poisson Distribution unit fixed?

Kind of
inspection No, varies Use u-Chart
variable?

Continuous Which spread Range Use X-bar and

method R-Chart
Variable preferred?
Standard Deviation
Use X-bar and
S-Chart