TS-0003413 Measurement Systems Analysis Standard

2014 Tesla Proprietary and Confidential

Measurement Systems

Analysis Standard

Tesla Motors, Inc.

TS 0003413

Revision 1.00

Tesla Motors Inc. Page: 2 of 12

Measurement Systems Analysis Standard TS-0003413

Revision 1.00

2014 Tesla Proprietary and Confidential UNCONTROLLED COPY Individuals using an uncontrolled copy are responsible for ensuring use of latest revision.

Contents

1.0 Introduction ..............................................................................................................................3

1.1 Scope .....................................................................................................................................3

1.2 Objective ...............................................................................................................................3

1.3 Dimensional Guidelines and References ..............................................................................3

2.0 Background ...............................................................................................................................4

2.1 Repeatability and Bias ..........................................................................................................4

3.0 Measurement Systems Analysis ...............................................................................................5

3.1 Properties of a good measurement System .........................................................................5

3.2 Testing Preparation ..............................................................................................................5

3.3 Selecting Sample Sizes ..........................................................................................................6

4.0 Gauge Repeatability and Reproducibility .................................................................................7

4.1 Analysis of Variance (ANOVA) Method ................................................................................7

4.2 Average Range Method ........................................................................................................7

4.3 Gauge Acceptability Criteria .................................................................................................9

5.0 Gauge Repeatability ...............................................................................................................10

5.1 GR Testing ...........................................................................................................................10

5.2 Analysis ...............................................................................................................................10

6.0 Works Cited ............................................................................................................................11

7.0 Revision History ......................................................................................................................12

Table of Figures

Figure 1 - Example of Repeatability and Bias (AIAG, 2010) .................................................................4

Figure 2- Impact of Number of Distinct Categories (ndc) of the Process Distribution on Control and

Analysis (AIAG, 2010) ..........................................................................................................................8

Figure 3 - GR&R Acceptability Criteria (AIAG, 2010) ...........................................................................9



Revision 1.00


1.0 Introduction

1.1 Scope

This document will focus on the methodology behind measurement system analysis.

Key concepts will be covered, including the preparation, analysis and actionable items of

measurement system analysis, specifically check fixtures and gauges that collect variable

data.

1.2 Objective

The objective of this document is to highlight the importance of Measurement

System Analysis (MSA), Gauge Repeatability and Reproducibility how to analyze a given

measurement system.

The end goal is to determine the amount of error in the measurement system and

asses the adequacy for product and process control.

1.3 Dimensional Guidelines and References

1.3.1 Tesla Motors, Inc. will interpret all dimensions on parts and assemblies

using the ASME-Y14.5 2009 Standard

1.3.2 Master source of all dimensional information is CATIA FT&A found within

the CATPart and CATProduct. FT&A will supersede all other formats

1.3.3 For information on inspection standards, please refer to TS-xxxxxx

1.3.4 For information on Teslas dimensional standards, please refer to

Dimensional Standards for Design, TS-0002978

1.3.5 For further reading or details on anything listed in this document, please

refer to the AIAG Measurement Systems Analysis Reference Manual, 4th

edition.

1.3.6 Gauge Repeatability (GR) Studies should utilize TS-0003544 for data

collection and analysis.

1.3.7 Gauge Repeatability and Reproducibility (GR&R) Studies should utilize TS-

0003552 for data collection and analysis.



Revision 1.00


2.0 Background

2.1 Repeatability and Bias

Repeatability and bias are a measure of accuracy and precision.

Repeatability is best defined as the ability to achieve the same results with a

repeated measurement. Bias is the shift of a grouping of results away from the

target. It is important to note that bias and repeatability are independent of each

other, and that controlling one of these sources of error does not guarantee the

control of the other (AIAG, 2010).

Figure 1 - Example of Repeatability and Bias (AIAG, 2010)



Revision 1.00


3.0 Measurement Systems Analysis

3.1 Properties of a good measurement System

The quality of a measurement system is usually determined solely by the statistical

properties of the data it produces over time. (AIAG, 2010) The following are fundamental

properties that will define a good measurement system:

3.1.1 Adequate discrimination and sensitivity is fundamental. Increments of

measure should be small relative the process variation or specification

limits, whichever is smaller. (AIAG, 2010) A good rule of thumb is the 10-to-

1 rule; the instrument discrimination should be at least one tenth of the

specification limits (or process variation).

3.1.2 The variability of a measurement system must be small compared to the

specification limits. (AIAG, 2010).

3.1.3 For process control, the variability of the measurement system should have

a small effective resolution when compared to manufacturing process

variation. Assess the measurement system to the 6-sigma process variation

and/or Total Variation from the MSA study (AIAG, 2010).

3.2 Testing Preparation

AIAG highlights the importance of planning and preparation prior to conducting a

measurement systems study. The following points should be reviewed ahead of any study.

(AIAG, 2010)

3.2.1 Understand the measurement approach Understand the tools and

equipment used to the measurement system in question.

3.2.2 Number of appraisers, sample parts, and repetitions There are a few

factors to consider when selecting sample sizes.

a) Criticality of dimensions Critical dimensions will require more parts

and trials. Ultimately, the degree of confidence required will drive the

parameters of the measurement system study.

b) Part Configuration Flexible parts may require more samples and trials

when compared to rigid parts.



Revision 1.00


c) Tesla Requirements Tesla may require deviation of standard process

parameters if there is a need for increased confidence in the

measurement system such as safety critical dimensions.

3.2.3 Chose appraisers and parts that represent the production process avoid

support from those who will not normally utilize the measurement system,

and avoid the use of special-build parts unless necessary (such as in Alpha

and Beta builds). Parts should be selected in a manner which will represent

the process by selecting parts over multiple builds, batches or tools.

3.2.4 The measurement instruments should be able to discriminate at least one-

tenth of the expected process variation (AIAG, 2010).

3.2.5 All measurements should be carried out following the defined

measurement procedure.

3.2.6 The study should be carried out in a random order within reason in

order to minimize any effect of drift through the study. The appraisers

should not be aware of which part is being measured, although the person

conducting the study should be tracking all the details.

3.2.7 The study should be managed and observed by someone who understands

the importance of conducting a reliable study.

3.3 Selecting Sample Sizes

The purpose of any measurement study is to understand the properties of the

measurement system.

3.3.1 A sufficient number of samples should be selected to cover the expected

operation range.

3.3.2 Must weigh in the risk of having defective material flow through the

operation. Higher risk will warrant a higher sample size.

3.3.3 For all new parts and assemblies, Tesla requires a standard 10-3-3 test; ten

parts, 3 appraisers and 3 trials.

3.3.4 For any re-certifications or minor changes to a part number, the GR&R can

be re-certified using a 5-3-3. Further details are covered in Section 3.2.2.



Revision 1.00


4.0 Gauge Repeatability and Reproducibility

Gauge Repeatability and Reproducibility, or GR&R, is the primary method of analyzing the

statistical properties of check fixtures or gauges that collect variable data. Tesla requires that the

GR&R is run using a 10-3-3 configuration, as per section 3.3. The GR&R can be performed using a

few different methods. Range method, Average and Range method, and the ANOVA method

(AIAG, 2010). The ANOVA method is Teslas preference because it includes the interaction of

operator to part, and is generally a more accurate methodology. Due to its wide spread use, Tesla

is also accepting the AIAG Average-Range methodology, covered in Section 4.2.

4.1 Analysis of Variance (ANOVA) Method

In the analysis of variance method, the variance can be decomposed into four

categories: Parts, Appraisers, the Interaction between parts and appraisers, and the

Repeatability of the gauge (AIAG, 2010). Note that randomization is an important factor to

the success of the ANOVA method See section 3.2.6.

4.2 Average Range Method

The Average Range methodology will calculate the percent of process variation of

the overall measurement system, including its main components; repeatability,

reproducibility and part variation.

Data is collected into the AIAG standard format for GR&R using percent tolerance

calculations at 6 Sigma. Using any basic computer software, the reports page can be

tabulated automatically, and the results will be displayed. It is important to ensure the

correct K constants are used when calculating EV, AV and PV. Once the data input has

been completed, we can now begin to analyze the final values. TS-0003552 is the only

accepted format for data collection and analysis for GR&R studies.

4.2.1 Equipment Variation EV This is the variation estimated to be the result

of the equipment performance, and also an index of the system

repeatability.

4.2.2 Appraiser Variation AV This is the variation estimated to be the result of

the appraiser performance, and also an index of the system reproducibility.

4.2.3 The GR&R is simply the sum of squares of the EV and AV, divided by the

tolerance rage. This is a good indication if your measurement system is

acceptable.



Revision 1.00


4.2.4 Number of Distinct Categories ndc is a measure of the number of

distinct categories that can be reliably distinguished by the measurement

system. The ndc is defined as the number of non-overlapping 97%

confidence intervals that will span the expected product variation (AIAG,

2010). The result should be greater than or equal to 5. See Figure 3.

Figure 2- Impact of Number of Distinct Categories (ndc) of the Process Distribution on Control and Analysis (AIAG, 2010)



Revision 1.00


4.3 Gauge Acceptability Criteria

Refer to the Table below for gauge acceptability criteria.

4.3.1 All gauges conform to Teslas TS-0003139, Gauge and Check Fixture

Standard

4.3.2 GR&R will be calculated using the percent tolerance calculations.

4.3.3 %GR&R results 10% or below are acceptable.

4.3.4 %GR&R results between 10%-30% require buy-off by all Tesla

representatives involved (Including Dimensional Engineering) See section

2.0. Fixture and inspection method modifications are likely.

4.3.5 GR&R results >30% are not acceptable. Corrections need to be made to the

fixture to improve the results and a new GR&R needs to be done. Fixture

changes need to be approved by Tesla Dimensional Engineering.

Figure 3 - GR&R Acceptability Criteria (AIAG, 2010)

4.3.6 Number of Discrete Categories should be greater than 5, See section 4.2.4

for more details. Departure from this requirement requires buyoff from

Tesla Motors Dimensional Engineering.



Revision 1.00


5.0 Gauge Repeatability

Gauge Repeatability (GR) can be used to analyze the variation in the measurement system

from only the gauge without the effect of the appraiser or part. This can be used as a

preliminary check for screening before shipping from gauge supplier or for pre-production

activities when part availability is minimal. The GR alone is not a sufficient test for gauge buy-off

or part acceptance.

5.1 GR Testing

A standard GR test will require 3 parts, 1 appraiser, and 10 trials (3-1-10). If a GR

test is used for preliminary buyoff or screening, testing will require 3 parts, 1 appraiser,

and 10 trials (3-1-10). Data should be collected using the Tesla GR Data Collection Sheet,

TS-0003544.

5.2 Analysis

Once the data has been collected, the analysis is simply the following equation:

% =6

100 =

1

1( )2

=1

When analyzing the 3-1-10 GR, the largest GR value from the group of 3 parts will

be used to represent the final outcome. Similar to the acceptability requirements for

GR&R, outlined in 4.3.5, the GR results must be below 10%.



Revision 1.00


6.0 Works Cited

AIAG. (2010). Measurement Systems Analysis (4th ed.). Southfield, MI: Automotive Industry

Action Group.



Revision 1.00


7.0 Revision History

Revision Level Date Description of Change

1.00 01/2015 Initial Release

Documents

TS-0003413 Measurement Systems Analysis Standard