Professional Documents
Culture Documents
System Analysis
Kevin B. Craner
Boise State University
October 6, 2003
Overview
Purpose of measurement system analysis (MSA)
Organizational uses of MSA
Measurement system errors
Accuracy and Precision of measurements
MSA process flow
Real world application
Exercise
Measurement System
Analysis (MSA)
aka Gauge R&R
A tool used to evaluate the statistical properties of
process measurement systems.
Purpose of MSA is to statistically verify that
current measurement systems provide:
Representative values of the characteristic being
measured
Unbiased results
Minimal variability
Organizational Uses
Mandatory requirement for QS 9000
certification.
Identify potential source of process
variation.
Minimize defects.
Increase product quality.
Measurement System Errors
Accuracy: difference between the observed
measurement and the actual measurement.
Process:
1. Determine which measurement system will
be studied.
2. Establish test procedure.
Preparation for Study
-Cont-
3. Establish the number of sample parts, the
number of repeated readings, and the
number of operators that will be used.
4. Choose operators and sample parts.
Evaluate Stability
Objective: evaluate measurement system
to determine if the system is in statistical
control.
Procedure:
1. Choose sample standards.
2. Measure sample standards three to five
times.
Evaluate Stability
-Cont-
3. Plot data on a x-bar and R chart.
Analysis:
1. Determine if process is in control.
2. If process is unstable determine and correct
the cause.
Evaluate Resolution
Objective: determine if the measurement
system can identify and differentiate
between small changes in the given
characteristic.
Process:
1. Choose a sample standard.
Evaluate Resolution
-Cont-
2. Measure the sample standard three to five
times.
Process:
1. Choose sample standards.
2. Measure sample standards 15 to 25 times using
the same measuring device, the same operator, and
the same setup.
Determine Accuracy
-Cont-
3. Calculate x-bar
4. Calculate bias
- Bias = Average Reference Value
5. Calculate the upper and lower 95%
confidence limit (CL).
Determine Accuracy
-Cont-
Analysis
Process:
1. Choose three to five sample standards that
cover the measurement space.
Evaluate Linearity
-Cont-
2. Measure sample standards 15 to 25 times.
3. Calculate the average of the readings.
4. Calculate bias.
5. Plot reference values on x-y graph.
6. Calculate slope of the linear regression line.
7. Calculate linearity and percent linearity.
8. Calculate R2.
Evaluate Linearity
-Cont-
Analysis
Process
1. Generate random order for operators and
parts to complete the run.
Determine Repeatability and
Reproducibility -Cont-
2. Repeat process for subsequent runs.
3. Have operators take measurements.
Analysis:
1. Plot data
2. Run ANOVA (analysis of variance) on data.
Determine Repeatability and
Reproducibility -Cont-
3. Calculate total variance.
4. Calculate % Contribution and determine if
acceptable.
5. Calculate % Contribution (R&R)
6. Calculate Process to Tolerance ratio (P/T)
for repeatability.
7. Determine if P/T is acceptable.
Real World Application
Case study performed by Tirthankar
Dasgupta and S.V.S.N Murthy, Indian
Statistical Institute, 2001.
- Gauge R&R study of automobile radiator
manufacturer.
- After studying four characteristics of
radiator components the following results
were obtained:
Table 1. Results of Preliminary
Study
Least
Characteristic Gauge Count TV %EV %AV %R&R
Filler neck hole dia. Vernier calipers 0.01 0.1107 35.77 72.88 81.12
Tube crown thickness Dial vernier calipers 0.02 0.1866 49.19 38.06 62.1
Tube width Dial vernier calipers 0.02 0.0365 10.71 43.83 46.87
Inlet hole dia. Vernier calipers 0.01 0.1057 86.56 20.09 88.83
Source: Dasgupta & Murthy, Total Quality Management, Vol. 12, No. 6, 2001
Real World Application
-Cont-
Any system having greater than 30% gauge
R&R is considered inadequate. As seen in
Table 1, all four characteristics %R&R is
inadequate.
Investigation of the measurement system
led to a subsequent reduction of %R&R in
three of the four characteristics to between
12% and 23%.
Real World Application
-Cont-
Further investigation of the fourth
characteristic, inlet hole diameter, led the
examiners to a manufacturing problem. The
team discovered high ovality in the inlet
hole, which was caused by the cutting tool.
The tool was modified to reduce ovality.
Real World Application
-Cont-
Benefits of the study
1. Reduced measurement variation.
2. Increased operator confidence regarding
their aptitude for conducting gauge R&R
studies.
3. Paved the way for further studies within the
firm.
An Exercise Calculating EV,
AV, R&R, and TV
Given: EV = 5.15(s0) , AV = 5.25(s1)
R&R = (EV2 + AV2)
TV = (EV2 + AV2 + PV2)
Where: s0 = gauge standard deviation = 0.05
s1 = true appraiser standard
deviation = 0.1
PV = part-to-part variation = 0.02
An Exercise Calculating EV,
AV, R&R, and TV