2. BASIC S.P.C.
WHAT IS S.P.C.?
STATISTICAL PROCESS CONTROL
IS THE USE OF STATISTICAL TECHNIQUES
(SUCH AS CONTROL CHARTS) TO
ANALYZE A PROCESS OR ITS OUTPUT.
S.P.C. HELPS MAINTAIN QUALITY WITHIN A
PROCESS BY:
Increasing customer satisfaction by reducing the amount of
variation, producing a more trouble free product .
Decreases scrap, rework, and inspection costs by
controlling the product.
Decreases operating costs by optimizing the frequency of
tool adjustments and changes.
Maximizes productivity by identifying and eliminating the
causes of out of control conditions.
Establishes a predictable and consistent level of quality.
Eliminates or reduces the need for receiving inspection by
the customer.
3. BASIC S.P.C.
All production processes have inherent variation. That is, no
two pieces produced are exactly alike if measured with
enough precision.
Variation can be categorized in terms of "common" and
"special" causes:
•Common cause variation occurs where a stable process
makes products that vary within a predictable range.
•Special cause variation results from unpredictable events
such as nonconforming raw material, a broken tool, or a
power sag.
SPC gives us the tools to measure the degree of both types of
variation with the goal being to eliminate special causes
altogether, and systematically attack common causes to reduce
them over time.
Control charts and capability studies are the main tools used to
describe processes graphically. Control charts are composed
of sampling results taken over time that are plotted as points
on special graphs. Different kinds of control charts
accommodate variable or attribute sampling.
4. BASIC S.P.C.
VARIABLE & ATTRIBUTE CHARACTERISTICS
A variable characteristic is a measurable feature such
as height, width, or weight.
Variable control charts are composed of two graphs. The
top graph monitors process statistical location. It
measures whether the process is adjusted
properly, comparing the calculated process average to
the print nominal or target value. The bottom graph
monitors process variation.
**********************************************************
An attribute characteristic is a countable feature such
as go/no-go, present/not present or number of defects.
Attribute control charts are composed of only one graph
that monitors lot-to-lot variations in terms of percent or
number nonconforming. While the target for
nonconformities is always zero, many unrefined
processes exhibit common-cause variation causing a
"predicted" nonconforming level above zero.
5. BASIC S.P.C.
BASIC STEPS OF S.P.C.
All tasks are to be performed during production
At predetermined intervals, select a subgroup of
sample products from the process (while it is running)
Measure the specified characteristic of each sample
Record the measurement values on a control chart
Calculate two values for each subgroup:
Average of the samples (X)
Range between the highest and lowest values (R)
Plot each subgroup value as a point on the
corresponding chart
Draw a line to connect each point to the previous point
on the chart
Analyze the patterns that develop as points are added
to the charts
6. BASIC S.P.C.
X – R CHART
The X & R chart is actually two charts, and is the most
commonly used method of tracking variable characteristics.
Only one characteristic can be recorded on each chart.
The X chart monitors subgroup averages
The R chart monitors subgroup ranges
A central line, which is a solid line across each
chart, represents the average of the subgroup values
On an X chart, it’s labeled X, and represents the
average of all the subgroup averages
On an R chart, it’s labeled R, and represents the
average of all of the subgroup ranges
Control limits, which appear as dashed lines across each
chart, represent the expected range of variation for the
subgroup values
The upper control limits is typically labeled UCL
The lower control limit is typically labeled LCL
The LCL on the R chart is often zero, and is
represented by the bottom edge of the graph
7. BASIC S.P.C.
DIFFERENCE BETWEEN CONTROL LIMITS AND
SPECIFICATION LIMITS
Specification limits define an allowable amount of variation
for a characteristic. These limits are determined when a product
is designed.
Control limits identify the expected range of variation for a
characteristic. These limits are based on the actual
measurements of the characteristic, as found during processing.
CALCULATING SUBGROUP VALUES
To find the X value for a subgroup:
• Add the values of all the sample measurements taken
• Divide this sum by the number of samples
To find the R value for a subgroup:
• Find the largest and smallest values in the subgroup
• Subtract the smaller value from the larger value
8. BASIC S.P.C.
1 2 3 4 5 6 7 8 9 10
38 31 30 32 32 33 33 33 33 30
35 31 30 33 34 38 34 33 36 35
DATA FROM EACH
34 34 32 33 37 31 36 36 35 37
SUBGROUP
34 31 29 32 37 33 31 35 34 32
29 32 32 37 35 33 30 36 31 29
TOTAL 170 159 153 167 175 168 164 173 169 163
AVERAGE ( X ) 34.0 31.8 30.6 33.4 35.0 33.6 32.8 34.6 33.8 32.6
RANGE ( R ) 9 3 3 5 5 7 6 3 5 8
37 UPPER CONTROL LIMIT
36
CHART OF 35
THE 34
AVERAGES 33 X
32
(X) 31
30
LOWER CONTROL LIMIT
29
1 2 3 4 5 6 7 8 9 10
SUBGROUPS
18
16
CHART OF 14
THE 12 UPPER CONTROL LIMIT
RANGES 10
8
(R) 6 R
4
2
1 2 3 4 5 6 7 8 9 10
SUBGROUPS
9. BASIC S.P.C.
PROCESS CAPABILITY AND PERFORMANCE
Process capability is a measure of the ability of a process to
produce products that meet required specifications. It is
measured by taking variable measurements over TIME from a
process that is statistically stable (only common cause variation
present). Process capability is typically reported as a measurable
referred to as the CpK value.
Cpk attempts to answer the question “will my process continue
to meet specifications in the long run?" Process capability
evaluation can only be done after the process is brought into
statistical control. The reason is simple: Cpk is a prediction based on
variation within a subgroup, and one can only predict something that
is stable. Think of the Cpk as the potential performance, or the
potential capability. Cpk is the CAPABILITY of your process if all
instability (special causes) were removed (or ignored).
Process performance is the actual state of the process at some
moment in time. In essence, a snap shot of the process now. In a
minute, hour, day, or week later it probably will be different. Process
performance is typically reported as a measurable referred to as the
PpK value.
Ppk attempts to answer the question "does my current
production sample meet specification?“ Ppk is how the process is
actually performing at the time you made the measurements. Think of
Ppk as being the actual PERFORMANCE of your
process, incorporating all observed variation within the subgroups.
10. STOGRAM 6
5
S.P.C.
4
3 BASIC
2
1
CAPABILITY31 32 33 34 HISTOGRAMS
29 30
ANALYSIS: 35 36 37 38
10
9 X
8 X
7 X X
STOGRAM 6 X X X X
5 X X X X X
4 X X X X X X X
3 X X X X X X X X X
HISTOGRAMS
2 X X X X X X X X X X
1 X X X X X X X X X X
29 30 31 32 33 34 35 36 37 38
10
9 X
8 X
7 X X
6 X X X X
5 X X X X X
4 X X X X X X X
3 X X X X X X X X X
2 X X X X X X X X X X
1 X X X X X X X X X X
29 30 31 32 33 34 35 36 37 38
VARIABLE DATA - NORMAL CURVE
1 2 3 4 5 6 7 8 9 10
38 31 30 32 32 33 33 33 33 30
35 31 30 33 34 38 34 33 36 35
DATA FROM EACH
34 34 32 33 37 31 36 36 35 37
SUBGROUP
34 31 29 32 37 33 31 35 34 32
29 32 32 37 35 33 30 36 31 29
TOTAL 170 159 153 167 175 168 164 173 169 163
AVERAGE ( X ) 34.0 31.8 30.6 33.4 35.0 33.6 32.8 34.6 33.8 32.6
RANGE ( R ) 9 3 3 5 5 7 6 3 5 8
11. BASIC S.P.C.
Cpk > 1.33 (Capable ) Cpk = 1 to 1.33 (Barely Capable)
This process should produce This process will produce
less than 64 PPM >64 PPM but < 2700 PPM
A Highly Capable Process : Voice of the A Barely Capable Process : Voice of the
Process < Specification Process = Customer Specification
********************************************
Cpk < 1 (Not Capable )
Cpk < 1 (Not Capable )
This process will also produce
This process will produce
more than 2700 PPM
more than 2700 PPM
A Non-Capable Process : Voice of the A Non-Capable Process : Voice of the
Process > Customer Specification Process > Customer Specification
12. BASIC S.P.C.
CONTROL CHART INTERPRETATION
NATURAL PATTERNS:
• Points remain within the control limits and vary randomly on both sides
of the central line
• Most points fall near the central line
• Some points fall closer to both control limits
• No points should fall beyond either control limit
•The process is:
• In statistical control
• Producing products within the expected limits of variation
• Operating as well as possible without adjustments
• Stable and predictable
• Variation is due to common causes – normal factors not easily identified
or eliminated from a process
• Represents a normal distribution
UNNATURAL PATTERNS:
• The process is:
• Out of statistical control
•Unstable and unpredictable
•Producing unexpected variation among products
• Variation may be due to assignable causes – unusual factors that can be
identified and eliminated from a process
•Common types are outliers, trends, runs or sudden changes in level
13. BASIC S.P.C.
CONTROL CHART INTERPRETATION
NORMAL DISTRIBUTION: Points vary randomly on either
side of the central line and remain within the control limits
OUT OF CONTROL CONDITION (OCC): One or more
points outside of the upper or lower control limits (requires a
user note and immediate corrective action)
14. BASIC S.P.C.
CONTROL CHART INTERPRETATION
TRENDS: A series of 6 or more consecutive points which
either ascend or descend across the chart; indicates a
problem with the process which must be addressed
RUNS: A series of 6 or more consecutive points which fall
either above or below the central line
