Schilling RoboticsHow to: Gauge Repeatability and Reproducability

August, 2012

This is an interactive Presentation: Click on the start button to begin!

The Big Picture

2

3

What is GR&R

• Gauge Repeatability and Reproducibility• Repeatable – One operator can achieve the same

measurement multiple times• Reproducible – Multiple operators can achieve the

same measurement. • Gauge – is an estimate• Gage - is an instrument used to measure a

magnitude or position

If at any time you need a definition click the info button and see if it is on the definition page

Why do we Care?

GR&R provides a quantitative way for the IQC department to validate our mechanical measuring systems and operators.

4

Test Breakdown

5

Home Screen

Using Minitab

Analyzing Results*Click on an image to access the information below it

Running The Study

Starting Your GR&R Test

It is important to understand what you are trying to accomplish before running a study.

CLICK ON THE ICON NEXT TO A TEST TO UNDERSTAND WHAT THAT PARTICULAR TEST ACCOMPLISHES

Testing the Capability of an Operator

Testing the Accuracy & Precision of a Measuring Device

Testing the Quality of Incoming Parts

6

Testing the Capability of an Operator

Purpose: This test is designed to show how well operators are using measuring devices.

End Result: It can reveal if operators need training or are sufficient on that particular

measuring device.

7

How to Set up a GR&R to Determine:The Capability of an Operator

8

The Capability of an Operator: Selecting Operators

Select the desired operators to be evaluated. Good choices include but are not limited to the following–New hires–All employees on new machinery–Trainers to ensure they are competent–Inspectors who differ from the average “accept/reject” rate–Must have at least 2 operators per study.

9

The Capability of an Operator:Selecting a Measuring Device

When selecting a measuring device make sure…

– To select a device commonly used by the operators involved

– The device is calibrated– Each operator uses the exact

same device– The devices operational functions

are not compromised before or during the testing period.

10

The Capability of an Operator:Selecting Parts

When picking parts make sure you

– Use parts that have tolerance ranges that match the measuring device you selected

– Use “Bad Parts”– Use completely different parts– Use both parts that the

operators have and haven't seen

11

Back to Tests

How to Set up a GR&R to Determine:The Accuracy & Precision of a

Measuring Device

12

Testing the Accuracy & Precision of a Measuring DevicePurpose: This test is designed to validate that a

measuring device is accurate & precise for a particular tolerance range.

End Result: Lead to re-calibration of a device or change in the tolerance range that the measuring device is allowed to inspect. Can also help discover machines that are not worth their cost.

13

The Accuracy & Precision of a Measuring DeviceSelecting Operators

Select operators who are competent on the selected measuring device.–Can use another GR&R to verify competency of the operators–Must have at least 2 operators per study

14

The Accuracy & Precision of a Measuring Device:Selecting a Measuring Device

Select the measuring device of your choice.

– Should be a device that has reason to be tested

15

The Accuracy & Precision of a Measuring Device:Selecting Parts

When picking parts make sure you– Use parts that have tolerance ranges that

match the measuring device you selected– Use “Bad Parts”– Use parts that cover a certain tolerance

range– Use both parts that the operators have and

haven't seen

16

Back to Tests

Running the Study

17

Guide Lines for an Effective Study • The Manager of the study should be present during every

trial in the study• Each trail must be conducted in similar environments• Study should be conducted in a RANDOMIZED order• Conduct Studies in a Non-bias fashion• Minimize error by defining the measuring objective

completely.• Do not reveal operators names when presenting you findings• Every study should have…

– At least 2 operators (3)– At least 5 parts (10)– At least 2 trails (3)

18

How to Set up a GR&R to Determine:The Quality of Incoming Parts

19

Testing the Quality of Incoming Parts

• Purpose: This test is equivalent to IQC inspection.

• End Result: Will determine if a part or group of parts is in or out of spec.

20

The Quality of Incoming Parts:Selecting Operators

Select the top operators on the selected measuring device.–Must have at least 3 operators per study–The more GOOD operators used the better the results

21

The Quality of Incoming Parts:Selecting a Measuring Device

When selecting a measuring device…

– Select a measuring device that is in scope of the parts tolerance range

– Should be a device that has been proven accurate and precise

22

The Quality of Incoming Parts:Selecting Parts

Use the incoming parts in question

23

Back to Tests

Pick Wisely!

24

How to Use Minitab to Run a GR&R Study

25

Open a GR&R Study Worksheet1. Open the program2. Stat menu quality tools Gauge Study Gauge R&R Study Worksheet

26

Open a GR&R Study Worksheet

27

• Now complete the pop up box with the following information.− Enter your number of parts (minimum of 5, typically 10+)− Enter your number of Operators (minimum of 2, typically 3)− Enter the number of replicates (minimum of 2, typically 3+)− Now Click Options

− Ensure that “randomize runs within operator” is checked and click “OK”− Click “Ok” to create the study

28

NOTE: There is no optimum number of parts, operators or replications to use. The more the study has of each, the more variation you can reveal but the more variation you introduce. It all depends on the system you are attempting to analyze

Editing GR&R Study Worksheet

29

Add a measurement column to the worksheet and print out individual copies for each operator for each trial.

NOW RUN YOUR STUDY & IPUT ALL THE DATA INTO MINITAB

Analyzing Results

30

31

Two Types of Data are Generated

Hard Data Graphical Data

32

Generating Your Own Data

Take Away

• P/T or %Tolerance: Is the measurement system good enough(precise) to measure against customer requirements.

• %R&R or %study variation: Is the measurement system good enough to detect changes in the process.

• %Contribution: How much variation each component of the study is providing

• P-T-P: Want to be as close to 100% as possible to show that the variation is coming from the parts not the measuring system.

33

Hard Data• The Numerical data that is generated

from the study• What you're looking for

– % Contribution

• x≤1% = Acceptable

• 1%≤x≤9% = Rework considered but acceptable

• x≥9% = Fail

– % Study Variation

• x≤10% = Acceptable

• 10%≤x≤30% = Rework considered but acceptable

• x≥30% = Fail

– Number of Distinct categories (NDC)

• NDC >5

34

Back to Analyzing Data

Graphical Data

35

Click on a gold Number to learn more about that graph!

Components of Variation

36

GOOD

BAD

The ideal graph has a low Gage R&R percentage, a larger Part-to-Part percentage and Repeat & Reprod percentages under 30%

R Chart By Operator

37

The ideal graph should have all points inside the UCL and LCL to be under control, If there are multiple points lying outside these limits it depicts that there is not enough consistency in the current measurement system

GOOD

BAD

38

GOOD

BAD

Xbar Chart by Operator

There should be a lot of part variation and all data on this plot should be outside the UCL and LCL, demonstrating part variation is easy to detect.

Variation by Part Sample

39

GOOD

BAD

The ideal chart should have a large variation between parts showing that the measurement system can detect different parts easily.

Measurement by Operators

40

GOOD

BAD

The ideal chart will have little variation within each operator [i.e. smaller gray box] and between all operators [i.e. horizontal straight line connecting the grey boxes]. This proves that the average is not effecting the final calculations.

41

Parts * Operator Interaction

GOOD

BAD

The ideal chart would have parallel lines meaning that there was no conflicting interactions proving the

measurement system is good. Back to Analyzing Data

Generating Your Own Data• One option is

Histograms• To generate: Graph Histogram

42

From here a pop up box will appear

asking you to select the data you

wish to graph.

43

Any Questions?

44

List of Machine Approved TolerancesGauge Accuracy Resolution

Minimum Tolerance for Use (4X)

Calipers (Digital) +/- .001 .0005 +/- .004Calipers (Dial) +/- .001 .001 +/- .004

Outside Micrometer (Std.) +/- .0001 .0001 +/- .0004

Outside Micrometer (Digital) +/- .0001 .0001 +/- .0004Groove Micrometer +/- .0004 .0001 +/- .0016Disc Micrometer +/- .0002 .0001 +/- .0008Thread Micrometer +/- .0002 .0001 +/- .004Tube Micrometer +/- .00015 .0001 +/- .00045Inside Micrometer +/- .001 .001 +/- .004Height Gauge (Digital) +/- .002 .0005 +/- .008Drop Indicator (Digital) +/- .00012 .00005 +/- .00048Drop Indication (Dial) +/-.0005 .001 +/-.002Test Indicator (0-15-0) +/-.00025 .0005 +/-.0001Bevel Protractor +/-5’ 5’ +/-20’

Digital Protractor +/-.1 deg +/-.1 deg +/-.4 deg

45

Source: Doc #130-1888