Dimensioning and Tolerancing Handbook Episode 3 Part 8 pdf

Initial analysis will focus on a positional tolerance in a nondiametral tolerance zone.Please note: A small sample size is used only out of convenience.. 4 Geometric controls @ MMC or LM

3 The American Society of Mechanical Engineers 1995 ASME Y14.5M-1994, Dimensioning and Tolerancing.

New York, NY: The American Society of Mechanical Engineers

Technol-25.1 Introduction

This chapter shows examples of calculating capabilities for a gage repeatability and reproducibility (GR&R)study on geometric tolerances, and identifies ambiguities as well as limitations in these calculations.Additionally, it shows tremendous areas of opportunity for future research and development in GR&Rcalculations due to past and still-current limitations in the variables considered when making these calcu-lations This chapter will define conditions not being accounted for in the calculations, therefore limitingthe measurement system’s capabilities

25.2 Standard GR&R Procedure

The following is a standard procedure used for calculating a GR&R that relates to geometric controls perASME Y14.5M-1994 Initial analysis will focus on a positional tolerance in a nondiametral tolerance zone.Please note: A small sample size is used only out of convenience Small sample sizes are strongly sup-ported when needing a quick “snap-shot” of a capability I do not, however, promote small sizes for in-depth analysis

Chapter

25

25-2 Chapter Twenty-five

Figure 25-1 Sample drawing #1

Table 25-1 GR&R Analysis Matrix

• Given 10 parts measured twice under the same conditions

• Same procedure

• Same machine

• Same person

• Resultant Values (R.V.) are to be shown in positional form (not just x or y displacement)

• Derive the range between runs for Part 1, Part 2, Part 10

• Sum the ranges and divide by 10 to derive the R

• Divide the R by a constant of 1.128, for sample/run size of 2 (rough estimate of sigma based on smallsample size)

• Multiply 3 × the estimate of sigma (3s) and divide by the positional tolerance allowed in the featurecontrol frame, then multiply × 100 (This derived value will represent the percentage of tolerance used

by the gage.)

The following data (Table 25-1) applies to the positional control of 0.2 mm, in relationship to datums

A primary and B secondary at regardless of feature size (RFS) as shown in Fig 25-1

Run #2

X displacement R.V.#2

RangeBetweenRV#1 &RV#2

2 ∆X +∆Y

Gage Repeatability and Reproducibility (GR&R) Calculations 25-3

Questions arise regarding these calculations and whether sigma should be multiplied by 3 or 6 Figs.25-2 and 25-3 are examples of tolerance zone differences, comparing a linear +/-0.1 mm tolerance to anondiametral position tolerance of 0.2 mm

Figure 25-2 Sample drawing #2 Figure 25-3 Sample drawing #3

25-4 Chapter Twenty-five

Based on the prior example, first impression might be to use only the linear displacement values tostay consistent with past and present Six Sigma conventions If only things were this simple, but they arenot In addition to the examples shown, there are many types of geometric callouts that require furtheranalysis of calculations to determine the most appropriate method of representing percentage of variablesgaging influence

The following is a beginning list of various types of geometric callouts that will need to be considered.1) Geometric controls @ RFS (diametral and nondiametral)

2) Geometric controls @ maximum material condition (MMC) or least material condition (LMC) (diametraland nondiametral)

3) Geometric controls @ MMC or LMC in relationship to datums that are features of size also defined atMMC or LMC

4) Geometric controls @ MMC or LMC with zero tolerance

Additional things not defined adequately deal with ranges for the following:

1) Features of size (lengths, widths, and diameters)

2) Linear plane to axis measurements

3) Axis (I.D.) to axis measurements

There are also questions as to which analysis methods to use (e.g., Western Electric, IBM, other).Also, what are the benefits, drawbacks and limitations of any of these methods?

Also, an acceptable method is needed to determine the bias of a measurement device with an able artifact, as well as a method to determine bias between devices Such a method must consider thefollowing:

accept-1) Sampling strategies

2) Spot size versus spacing versus sampling effects on a given feature

3) Replication of test (time versus environmental)

4) Confidence intervals

5) Truth (conformance to ASME Y14.5M-1994 and ASME Y14.5.1M-1994)

Note: For all geometric controls, the tolerance defined in the feature control frame is a “total ance,” of which the targeted value is “always” zero (0), and the upper control limit is always equal to thetotal tolerance defined (unless bonus tolerance is gained due to MMC or LMC on the considered feature).For geometric controls, such as the one shown in Fig 25-4, the 5 mm+/-0.2 mm diameter is positionedwithin a diametral tolerance zone of 0.02 mm at its maximum material condition, in relationship to datums A(primary), B (secondary), and C (tertiary) The following analysis is proposed:

toler-Gage Repeatability and Reproducibility (GR&R) Calculations 25-5

Figure 25-4 Sample drawing #4

The example shown in Fig 25-3 was for a nondiametral positional tolerance The example in Fig 25-4

is a diametral positional tolerance If this tolerance were defined at RFS rather than MMC, the procedurewould be identical to the one shown in support of Fig 25-3 The exception would be two additionalcolumns to represent the y-axis displacement from nominal In the example shown in Fig 25-4, the 0.02 mmdiametral tolerance zone applies only when the diameter of 5 mm is at its MMC size (4.8 mm) As it changes

in size toward its LMC size (5.2 mm), bonus tolerance is gained, as shown in the following matrix

Table 25-2 Bonus tolerance gained due to considered feature size

Feature of Size

∅5 +/- 0.2

Allowable PositionTol

con-25-6 Chapter Twenty-five

Table 25-3 Analysis Matrix

Gage Repeatability and Reproducibility (GR&R) Calculations 25-7

25.3 Summary

This chapter defined opportunities that will spur future research activities and should have made clearmany of the steps needed to determine a measurement system capability along with the reasons for strictand aggressive controls Discussions have started in 1998 within standards committees and universities

to concentrate resources to research and develop standards, technical reports, and other documentation

to further advance these analysis methods

Di-3 The American Society of Mechanical Engineers 1995 ASME Y14.5M-1994, Dimensioning and Tolerancing.

New York, New York: The American Society of Mechanical Engineers

P • A • R • T • 8

THE FUTURE

I thank the contributors for their wisdom and insight I look forward to seeing how these predictionsunfold.

Paul Drake

Timothy V Bogard

President, Sigmetrix

Dallas, Texas

The Future of Dimensional Management

Dimensional management as a methodology will continue to gain in acceptance with the more cated companies, where high volume and high complexity exist in the product lines The concept ofdimensional management will be of interest in other types of companies where low volume and lowcomplexity exists, but the cost of implementation in terms of training and process change will be the majorbarrier

sophisti-The Future of Geometric Dimensioning and Tolerancing (GD&T)

GD&T will continue to gain acceptance The standard(s) will need to continue to evolve to (1) eliminateambiguity, (2) improve assembly level tolerance definitions, and (3) be further consolidated to simplify theconcepts for more practical usage

Chapter

26

26-2 Chapter Twenty-six

The Future of Standards

Standards in the area of geometric definitions, like STEP (STandard for the Exchange of Product), are

critical to the long-term interoperability required by companies as they migrate across computer aideddesign systems, and further integrate with supplier base and customers There will continue to be empha-

sis on STEP like compliance by product developers and CAD/CAE (Computer Aided Design/Computer Aided Manufacturing) tools providers, to allow for flexibility and ease of use.

The Future of Tolerancing in Academics

More universities are already developing courses and research expertise in the area of tolerancing Better

alliances between industry and academia will need to be forged, like ADCATS (Association for the Development for Computer-Aided Tolerancing Systems) at BYU (Brigham Young University), to guaran-

tee the transfer of research to the industry If the research does not turn into easy-to-adopt concepts,methods and technologies, then the interest by industry in supporting academia will wain

The Future of Tolerancing in Business

As Six Sigma type initiatives continue to broaden and become the critical differentiator in many nies, tolerance analysis will elevate to the same level of importance as reliability and warranty analysis forall companies As ease of use continues to improve, the adoption of tolerance optimization techniques willproliferate in all areas of system design

compa-The Future of Software Tools

Software tools will go through a consolidation process whereby the basic analysis is a natural part of thedesign capture process Requirements flowdown, surface-based modeling and analysis, and partproducibility will become natural to the engineer, as the software tools providers continue to bury theprocess of tolerance optimization continuously in the design through manufacture process Basically,ease of use will dominate the tools suppliers agenda until the tolerancing process is virtually undetect-able

Kenneth W Chase, Ph.D.

Mechanical Engineering Department

Brigham Young University

Provo, Utah

Future of Tolerance Analysis

It is a pleasure to address the question: “What is the future of tolerance analysis?” It is a subject aboutwhich I have strong feelings I first began teaching a course in Design for Manufacture after returningfrom two summers working for John Deere in 1980 Two gray-haired engineers there, who were brothers,one a designer and the other a manufacturing engineer, persuaded me that mechanical engineers shouldinclude manufacturing considerations in their designs They spent a lot of time with me, “filling in thegaps in my education.”

I began to see that tolerance analysis was the vehicle to bring design and manufacturing together.Using a common mathematical model that combines the performance requirements of the designer withthe process requirements of the manufacturer provides a quantitative tool for estimating the effects eachhas upon the other It truly promotes the concept of Concurrent Engineering

2 Education and training.

As with other quality improvement programs, everyone involved must be educated about the rolethat tolerance analysis will play in the product development cycle and its expected benefits Manage-ment, design, production—all must catch the vision

The most challenging aspect is the fact that there is no established user base, no established lum, and there are no established procedures to guide us in implementing this new tool It is much easierfor a company to begin using an established CAD application, such as finite element analysis There aremany successful examples they can emulate But tolerance analysis is still in its infancy

curricu-The procedures for performing a finite element analysis are well established curricu-There are many lished examples Structural analysis departments are found in most big companies You can hire anexperienced person to help set up a program in your company But, this is not yet true for toleranceanalysis

pub-You can’t even hire the capability you need fresh out of school, because tolerance analysis is notfound in the curriculum of our engineering and technology schools Will it be there eventually? It is hard

to say The curriculum of our schools is under constant pressure Most schools have reduced the number

of hours required for graduation, while increasing the nontechnical requirements You can’t push ance analysis in, without pushing something else out

toler-For the time being, industry must expect to shoulder the burden of building the expertise they needwithin their own ranks Training seminars and consultants will be needed to assist in this effort

Unresolved Issues

Among the principal issues that must be resolved before CAD-based tolerance analysis is widely adopted:

1 The relationship to GD&T must be resolved.

There are many misconceptions about the application of GD&T standards to assembly toleranceanalysis How do MMC or RFS apply to a tolerance stackup? How about bonus tolerances? Aregeometric variations applied differently in a statistical analysis versus worst case? If a form tolerance is

26-4 Chapter Twenty-six

applied to a feature of size, should two variation sources be included in the tolerance stackup? Do the sizevariations include the surface variations, or do they represent two independent sources of variation?Most of the misconceptions arise from a lack of understanding of the fundamental principles uponwhich the GD&T standards and assembly tolerance analysis are based We also need to get a clearconcept of the difference between a specified tolerance and a measured or predicted variation

2 New standards for assembly variation are needed.

There are no standards for computing tolerance stackup and variation propagation in assemblies.ASME Y14.5 has only recently acknowledged the existence of statistical stackup analysis How it is to bedone is still open-ended This writer strongly feels that there should be a new set of symbols to differen-tiate an assembly tolerance limit involving multiple parts and a component tolerance limit applied to asingle part

3 Better data on process variations are needed.

The assembly variations predicted by tolerance analysis are only as accurate as the process variationdata entered into the analysis model However, there is very little published data describing processvariations and the cost associated with specified tolerance limits If you wait until the parts are made, someasured variations can be used in the model, you will lose one of the major benefits of tolerance analysis

In the design stage of a new product, tolerance analysis serves as a virtual prototype for predicting the

effects of manufacturing variations before the parts are made To fully realize this benefit, we simply musthave an extensive database, which characterizes process variations over a wide range of conditions andmaterials

4 Realistic expectations.

Over the years I have worked to involve industries and CAD vendors in the development of based tolerancing tools A number of companies have given enthusiastic support I have, however, beenturned away by several companies who have said in effect: “Come back when you have a finishedproduct.” Others seem to be waiting for “push-button tolerance analysis” that will require no understand-ing of variation and no decision-making skills

CAD-A state-of-the-art CCAD-AD tool cannot be developed without substantial resources and talent It needsbroad support from the CAD vendors and the end-users in industry CAD systems will require basicchanges in data structure to accommodate variation definitions CAD vendors must adopt standard userinterface tools and allow third-party access to possibly proprietary internal representations

Industry will need to take a more active role in guiding the CAD application development and oughly testing the resulting software products Industry must also develop an infrastructure for absorb-ing and implementing CAD-based tools into their product life cycle Until industries learn how to applytolerance analysis to their own enterprises, they will not be able to effectively influence its development