ENGM 620: Quality Management

Session 8 – 30 October 2012

• Process Capability
 Outline
• Process Capability
– Natural Tolerance Limits
– Histogram and Normal Probability Plot
• Process Capability Indices
– Cp
– Cpk
– Cpm & Cpkm
• Measurement System Capability
– Using Control Charts
– Using Factorial Experiment Design (ANOVA)
• Hands On Measurement System Capability Study
Process Capability - Timing
Process Capability Analysis is performed Improving Process Capability and
when there are NO special causes of Performance
variability present – ie. when the process is in
a state of statistical control, as illustrated at Continually Improve the
this point. System

Characterize Stable Process Capability

Head Off Shifts in Location,

Time Spread

Identify Special Causes - Bad

(Remove)
Identify Special Causes - Good
(Incorporate)
Reduce Variability

Center the Process

LSL 0 USL
 Natural Tolerance Limits
• The natural tolerance limits assume:
– The process is well-modeled by the Normal
Distribution
– Three sigma is an acceptable proportion of the
process to yield
• The Upper and Lower Natural Tolerance
Limits are derived from:
– The process mean () and
– The process standard deviation ()
• Equations:
UNTL    3
LNTL    3
Natural Tolerance Limits

  1 :68.26% of the total area

  2 :95.46% of the total area
  3 :99.73% of the total area

-3 - + +3
or -2  +2 or
LNTL UNTL

The Natural Tolerance Limits cover 99.73% of the process output

 Process Capability Indices
• Cp:
– Measures the potential capability of the
current process - if the process were centered
within the product specifications
– Two-sided Limits:
USL  LSL
Cp 
– One-sided Limit: 6

USL     LSL
Cpu  Cpl 
3 3
 Process Capability Ratio Note
• There are many ways we can estimate the
capability of our process
• If σ is unknown, we can replace it with one
of the following estimates:
– The sample standard deviation S
– R-bar / d2
 Process Capability Indices
• Cpk:
– Measures actual capability of current process
- at its’ current location with respect to product
specifications
– Formula:
C pk  min( C pu , C pl )
Where:
USL     LSL
Cpu  Cpl 
3 3
 Process Capability Indices
• Regarding Cp and Cpk:
– Both assume that the process is Normally
distributed
– Both assume that the process is in Statistical
Control
– When they are equal to each other, the
process is perfectly centered
– Both are pretty common reporting ratios
among vendors and purchasers
 Process Capability Indices
• Two very different processes can have
identical Cpk values, though:
– because spread and location interact!

LSL USL
PCR and an Off-Center Process
• CPK = min (CPU, CPL)

• Generally, if CP = CPK, then the process is

centered at the midpoint of the
specifications
• If CP ≠ CPK, then the process is off-center
 Comparison of Variances

– The second types of comparison are those that compare the

spread of two distributions. To do this:
• Compute the ratio of the two variances, and then compare the ratio
to one of two known distributions as a check to see if the magnitude
of that ratio is sufficiently unlikely for the distribution.

Definitely Probably Probably NOT Definitely NOT

Different Different Different Different

• The assumption that the data come from Normal distributions is very
important. Assess how normally data are distributed prior to
conducting either test.
 Process Capability Indices

• Cpm:
– Measures the current capability of the
process - using the process target center
point within the product specifications in
the calculation USL  LSL
Cpm 
– Formula:
6  2  (   T )2
1
Where target T is:T  (USL  LSL )
2
 Process Capability Indices

• Cpkm:
– Similar to Cpm - just more sensitive to
departures from the process target center
point
– Not really in very common use
C pk
– Formula: C pkm 
2
T
1  
  
 Measurement System
Capability
• Examines the relative variability in the
product and measurement systems,
together
– Total variation is the result of
• Product variation
• Gage variation
• Operator variation gaging system variation
• Random variation
 total   product   gage
2 2 2
Measurement System Analysis
• Measurement system can be assessed by
– X-bar and R-Charts
• Using a single part as the rational subgroup
• Is easy to visualize
• Requires alternate interpretation of the control
charts
– Designed Experiments
• Using Analysis of Variance
• Allows assessment of part x operator interactions
• Is statistically complex to compute & analyze
 X-Bar & R-Chart Method
• Have each operator measure the same part twice -
so the part becomes the rational sample unit
– Parts should be representative of those to be measured
• Use a sample of 20 - 25 parts
– Use a representative set of operators
• Either collect data from every operator, or
• Randomly select from the set of operators
– Collect data under representative conditions
• Carefully specify and control the conditions for measurement
• Randomly sequence the combination of parts and operators
• Preserve the time-order of the collected data & note observations
 X-Bar & R-Chart Method
• If each operator measures the same part
twice:
– Variation between samples is plotted on the X-
Chart
• Out of control points indicate success in identifying
differences between parts
– Variation within samples is plotted on the R-Chart
• Centerline of R-Chart is the magnitude of the gage
variation
• Out of control points indicate excessive operator to
operator variation (fix with training?)
 X-Bar & R-Chart Method

Out of control points indicate Out of control points indicate

ability to distinguish between inability of operators to use
product samples (Good) gaging system (Bad)

UCL

UCL
x
LCL
R

LCL

Sample Number Sample Number

X-Bar Control Chart R - Control Chart

 X-Bar & R-Chart Method

• Precision to Tolerance Ratio (P/T):

– “Rule of Ten”:
• The measurement device should be at least ten
times more accurate than the smallest measurement
– Calculations: and
 gage 
R P 6 gage
d2 
– Interpretation: T USL  LSL
• Resulting ratio should be 0.10 or smaller if the gage
is truly capable
X-Bar & R-Chart Method: R & R

• Repeatability:
– Inherent precision of the gage
• Reproducibility:
– Variability of the gage under differing
conditions
• Environment
• Operator
• Time …
 2
gage  2
repeatability  2
reproducability
X-Bar & R-Chart Method: R &
R
• Process is the same as before (20 - 25 parts, …):
– But we estimate the Repeatability from the Range
Mean computed across all the operators and all
parts:
R
 2
repeatabil ity 
d2
– And we estimate the Reproducibility from the
Range of variability across all operators for each
individual part:
R
 reproducab
2
ility 
x

d2