Statistical Process Control
Douglas M. Stewart, Ph.D. The Anderson Schools of Management The University of New Mexico
Quality Control (QC)
Control the activity of ensuring conformance to requirements and taking corrective action when necessary to correct problems Importance
Daily management of processes Prerequisite to longer-term improvements
Designing the QC System
Quality Policy and Quality Manual
Contract management, design control and purchasing Process control, inspection and testing Corrective action and continual improvement Controlling inspection, measuring and test equipment (metrology, measurement system analysis and calibration) Records, documentation and audits
Example of QC: HACCP System
1. Hazard analysis 2. Critical control points 3. Preventive measures with critical limits for each control point 4. Procedures to monitor the critical control points 5. Corrective actions when critical limits are not met 6. Verification procedures 7. Effective record keeping and documentation
Inspection/Testing Points
Receiving inspection In-process inspection Final inspection
Receiving Inspection
Spot check procedures 100 percent inspection Acceptance sampling
Acceptance Sampling
Lot received for inspection Sample selected and analyzed Results compared with acceptance criteria Accept the lot Send to production or to customer Reject the lot Decide on disposition 7
Pros and Cons of Acceptance Sampling
Arguments for:
Provides an assessment of risk Inexpensive and suited for destructive testing Requires less time than other approaches Requires less handling Reduces inspector fatigue
Arguments against:
Does not make sense for stable processes Only detects poor quality; does not help to prevent it Is non-value-added Does not help suppliers improve
In-Process Inspection
What to inspect?
Key quality characteristics that are related to cost or quality (customer requirements)
Where to inspect?
Key processes, especially high-cost and value-added
How much to inspect?
All, nothing, or a sample
9
Economic Model
C1 = cost of inspection and removal of nonconforming item C2 = cost of repair p = true fraction nonconforming Breakeven Analysis: p*C2 = C1 If p > C1 / C2 , use 100% inspection If p < C1 / C2 , do nothing
10
Human Factors in Inspection complexity defect rate repeated inspections inspection rate
Inspection should never be a means of assuring quality. The purpose of inspection should be to gather information to understand and improve the processes that produce products and services.
Gauges and Measuring Instruments
Variable gauges Fixed gauges Coordinate measuring machine Vision systems
12
Examples of Gauges
Metrology - Science of Measurement
Accuracy - closeness of agreement between an observed value and a standard Precision - closeness of agreement between randomly selected individual measurements
Repeatability and Reproducibility
Repeatability (equipment variation) variation in multiple measurements by an individual using the same instrument. Reproducibility (operator variation) variation in the same measuring instrument used by different individuals
Repeatability and Reproducibility Studies
Quantify and evaluate the capability of a measurement system
Select m operators and n parts Calibrate the measuring instrument Randomly measure each part by each operator for r trials Compute key statistics to quantify repeatability and reproducibility
Reliability and Reproducibility Studies(2)
Measurement (M) made by Operators (i from 1 to m) on Parts (j from 1 to n) in Trials (k from 1 to r) M ijk j k average for each operator xi = nr xD = max( xi ) min ( xi ) difference (range) of operator averages
i i
R ij = max ( M ijk ) min ( M ijk ) range for each part for each operator
k k
Rij j average range for each operator Ri = n Ri average range of all R = i m
Control limit of ranges Rij = D4 R for randomness of errors.
Reliability and Reproducibility Studies(3)
K1 is a constant tied to # of trials
Use number trials (r) for n in table. Check Repeatability or Equipment Variation EV = K1 R Reproducibility or operator (appraisal) variation EV 2 AV = ( K 2 xD ) nr K 2 is a constant tied to # of operators Repeatability and Reproducibility
2
R&R =
( EV ) 2 + ( AV ) 2
Results are in actual units measured. Customary to express as percentages. Under 10% - Acceptable 10 - 30% - ? based on importance and repair cost Over 30% - Unacceptable
R&R Constants
Number of Trials K1 Number of Operators K2 2 3 4 5
4.56 3.05 2.50 2.21 2 3 4 5
3.65 2.70 2.30 2.08
R&R Evaluation
Under 10% error - OK 10-30% error - may be OK over 30% error - unacceptable
R&R Example
R&R Study is to be conducted on a gauge being used to measure the thickness of a gasket having specification of 0.50 to 1.00 mm. We have three operators, each taking measurement on 10 parts in 2 separate trials.
x1 = 0.830 x2 = 0.774 x3 = 0.829
R1 = 0.037 R2 = 0.034 R3 = 0.017
Calibration
Calibration - comparing a measurement device or system to one having a known relationship to national standards Traceability to national standards maintained by NIST, National Institute of Standards and Technology
Statistical Process Control (SPC)
A methodology for monitoring a process to identify special causes of variation and signal the need to take corrective action when appropriate SPC relies on control charts
24
Common Causes
Special Causes
Histograms do not take into account changes over time.
Control charts can tell us when a process changes
Control Chart Applications
Establish state of statistical control Monitor a process and signal when it goes out of control Determine process capability
27
Commonly Used Control Charts
Variables data
x-bar and R-charts x-bar and s-charts Charts for individuals (x-charts)
Attribute data
For defectives (p-chart, np-chart) For defects (c-chart, u-chart)
28
Developing Control Charts
1. Prepare
Choose measurement Determine how to collect data, sample size, and frequency of sampling Set up an initial control chart
2. Collect Data
Record data Calculate appropriate statistics Plot statistics on chart
Next Steps
3. Determine trial control limits
Center line (process average) Compute UCL, LCL
3. Analyze and interpret results
Determine if in control Eliminate out-of-control points Recompute control limits as necessary
Typical Out-of-Control Patterns
Point outside control limits Sudden shift in process average Cycles Trends Hugging the center line Hugging the control limits Instability
36
Shift in Process Average
Identifying Potential Shifts
Cycles
Trend
Final Steps
5. Use as a problem-solving tool
Continue to collect and plot data Take corrective action when necessary
5. Compute process capability
Process Capability
Capability Indices
UTL LTL 6 if C p 1 is defined as capable (1.5 more often the minimum) Cp = Example : Part specification is 10.75mm .25mm = 0.0868mm
11.00 10.50 Cp = = 0.96 6 0.0868
UTL 3 LTL C pl = 3 C pk = min ( C pl , C pu ) C pu =
Process Capability (2)
C pu =
C pl =
2 T
11.0 10.7171 = 1.086 3 0.0868
10.7171 10.5 = 0.834 3 0.0868
C pk = C p (1 K ) where K =
Tolerance Example : same as above, but assume process is centered at 10.7171mm
C pm = 1+
C pm = 1+
Cp
( T )
2
0.960
T is the Target
(10.7171 10.75)
0.8682
= 0.8977
Capability Versus Control
Control Capability
Capable Not Capable In Control
IDEAL
Out of Control
44
Process Capability Calculations
Excel Template
Special Variables Control Charts
x-bar and s charts x-chart for individuals
Charts for Attributes
Fraction nonconforming (p-chart)
Fixed sample size Variable sample size
np-chart for number nonconforming Charts for defects
c-chart u-chart
Control Chart Selection
Quality Characteristic variable
defective n>1? yes n>=10 or no computer? yes x and s x and R no x and MR constant sample size? no p-chart with variable sample size yes p or np constant sampling unit? yes c no u
attribute
defect
64
Control Chart Design Issues
Basis for sampling Sample size Frequency of sampling Location of control limits
65
Pre-Control
LTL Red Zone UTL Red Zone
Green Zone
nominal value
Yellow Zones
67
SPC Implementation Requirements
Top management commitment Project champion Initial workable project Employee education and training Accurate measurement system
68