QMT Features: November 2013
Enabling the promises of GD&T
Understanding GD&T is vital for managing the risks involved in machine part design, manufacturing, inspection and assembly, and for getting it right first time. By Bill Tandler, president, Multimetrics Inc.

For some people, the acronym GD&T stands for “Grim, Depressing & Troublesome”, whereas of course it stands for “Geometric Dimensioning & Tolerancing”.  Why is it grim, depressing and troublesome for so many people?  Before we go there, let’s get a handle on what GD&T is all about.

In fact, GD&T is a symbolic language with which to 1) research, 2) refine and 3) ultimately “encode” the functions of each feature of a machine part by imposing permissible limits of imperfection in order to 4) guarantee assembly and operation.  In greater detail, its many contributions are:

1.    to encourage mechanical designers to think about the functions of the features of their parts and alter/refine their geometry to maximize their power to fulfill those functions early in the product development cycle,
2.    to enable mechanical designers to specify precisely defined, permissible limits of imperfection for each feature in order to guarantee performance of their functions in fault tolerant ways,
3.    to enable fully functional tolerance stack-up analysis with which to assess and refine the GD&T code imposed on parts,
4.    to guarantee the assembly and operation of parts prior to model release,
5.    to maximize production output by specifying manufacturing objectives so unambiguously as to make them totally independent of “tribal understandings”, and
6.    to enable full automation of coordinate metrology processes – again, independent of “tribal understandings” - and guarantee function based a) accept/reject decisions and b) manufacturing process feedback.

CAD (Computer Aided Design), and GAD (GD&T Aided Design) are the two equally important halves of the science of machine part geometry specification.  CAD provides the means to generate, manipulate and communicate the geometry of machine parts.  GAD on the other hand, serves to encode their functions by specifying permissible limits of imperfection.  CAD without GAD is very “BAD”, because it represents just half the story, and BAD GAD – “decorative” GD&T - is even worse, because it presents a false story.  Both alternatives lead to bumpy new product ramp up cycles, costly remanufacturing and highly questionable inspection results, all based on tribal understandings.  Only GOOD GAD is of any value.  In short, GD&T is a highly sophisticated “encodable” and “decodable” symbolic language for managing the risks involved in machine part design, manufacturing, inspection and assembly, and getting it right the first time.

What makes GD&T so “Grim, Depressing & Troublesome”?
GD&T is used broadly in aerospace, automotive and medical device manufacturing, but unfortunately largely to “decorate” drawings for later “interpretation” rather than to “encode” feature functions for unambiguous “decoding” in manufacturing and quality assurance.  Why is proper use of GD&T so difficult?

1.    The symbolic language of GD&T is extensive and complex, as it  must be in order to deal effectively with the complexity of the geometry and interactions of real parts.  As a result, mastering it takes time, strong determination and very special skills in the practical worlds of manufacturing, metrology, assembly and operation.
2.    Most universities have failed to incorporate strong GD&T training programs in their curriculums, leaving students to stumble upon it and struggle to master it in under the impact of the tribal understandings in the environments in which they work.
3.    Furthermore, the current American and International standards which define GD&T often add to the difficulty of using it due to instances of unclear terminology, incomplete definitions, failure to eliminate unnecessary tools and add missing tools, and most importantly, failure to clearly state and list the many rules on which definitive “encoding” and “decoding” depend.
4.    Finally, most GD&T is currently applied in the 2D drawing world of CAD, in which there can be no intelligent interaction between the computer and the design engineer.  For example, how can a computer forbid the use of a Material Condition modifier (S), (M) or (L) when Parallelism is applied to a planar surface, and require one when applied to the mid-plane of a slot, when there is no way to recognize the feature type in the 2D world?  Only operation in the 3D CAD world can enable “smart” guidance for the “encoding” process, and “smart” assessment of existing “code”.

What are the impacts of “Decorative” GD&T?
As opposed to functional GD&T,
1.    “Decorative” GD&T requires “interpretation”, leading to costly, time consuming debates between design, manufacturing and quality assurance team members, only after problems have surfaced.
2.    “Decorative” GD&T fails to define manufacturing objectives precisely, requiring contract suppliers to discover the actual objectives through hit and miss experiments, which, once discovered, often remain guarded secrets, making switching from one supplier to another a dangerous and expensive business.
3.    “Decorative” GD&T fails to define precise coordinate metrology objectives and processes, leading 1) to frequent acceptance of non-functional parts as well as rejection of functional ones, 2) to vastly different inspection reports from different suppliers based on different “tribal understandings”, and 3) to confused and therefore costly efforts to rectify manufacturing processes.

Examples of three worlds:
Given the flat edged vacuum flange illustrated and described below, let’s look at the impact 1) of “classical” dimensioning and tolerancing (CD&T), 2) of “decorative” GD&T and 3) of functional, syntactically correct GD&T.

Functional objectives of the part
1.    One face of the flange shall be flat within 0.01mm to create a reliable vacuum seal with an O-ring in a mating flange.
2.    The opposing face shall be located within 1mm but be parallel to the first within 0.2mm.
3.    The bolt hole pattern shall be centered in the flange within Ø1mm, and the bores mutually located within Ø0.2, to serve as mating part  locating features.
4.    The orientation and location of the flat edge on the flange shall be controlled within 0.2mm relative to the bolt hole pattern to serve to reliably align and locate another mating part.    

“Classical Dimensioning & Tolerancing”
1.    The classical drawing fails to make clear whether the location of the flat edge on the flange is to be controlled relative to the bolt hole pattern or to the OD of the flange.
2.    Lacking a Flatness tool, the form of the mating surface of the flange remains uncontrolled.
3.    Given that the part has two symmetry planes, it is impossible to clearly differentiate between symmetric features, thus complicating manufacturing process feedback and assembly.
4.    The “pie shaped” tolerance zones defined for the axes of the bores are totally non-functional.
5.    Definitive coordinate metrology is impossible.   

“Decorative GD&T”
1.    The symmetry problem remains, destroying the potential value of much of the rest of the code.
2.    The feature labelled Datum Feature A is not a mating feature and is therefore totally non-functional as the primary  Datum Feature  of this part as well as for coordinate metrology (the “aspect ratio problem”).
3.    The Parallelism tool applied to the flat edge does not impose the required location control in spite of the presence of the 80mm basic dimension which serves no purpose.
4.    All the material condition and boundary modifiers (M) need to be replaced by (S) in order for the code to begin to represent the actual functions of the controlled features. They are all locating features.
5.    The applicable GD&T standard has not been specified, making the controls meaningless.   

“Functional, Syntactically Correct, GD&T”
1.    The symmetry of the part has been broken to provide fully functional, unambiguous feature controls.
2.    The planar surface of the flange which engages the mating flange to constrain pitch, yaw and one degree of translational freedom is properly specified as the primary Datum Feature.
3.    The bolt hole pattern which aligns and locates this flange relative to the mating flange is properly specified as the secondary Datum Feature and the modifier (S) has been chosen to encode the “centering” function.
4.    The flat edge has been properly located relative to planar surface A and bolt hole pattern B using the Datum Feature modifier (S) to encode the centering function.
5.    The applicable GD&T standard has been identified.
6.    The axes of Datum Reference Frame [A,B] have been uniquely identified to impose definitive requirements for coordinate metrology reporting.
7.    The remaining requirements have been imposed in a maximally fault tolerant manner.
(Note:  The Modifier (S) has been used explicitly for clarity.)
What can we do to enable Functional, Syntactically Correct GD&T from the outset?
In other words, make GD&T “Grand, Delightful and Tantalizing”!
1.    It is time for an intelligent – i.e. SMART – 3D CAD based “GD&T Encoding & Decoding Engine” to take the helm, namely for an effective MBD (Model Based Definition) system to become available.  One would expect such an “engine” 1) to know all the rules and best practices of each widely used GD&T standard, 2) to be manipulable to fit individual corporate best practice criteria, and 3) to engage mechanical designers in supportive ways with explanations for pro and con recommendations and alternate paths through the “encoding” jungle, thereby turning that jungle into an English garden.
2.    It is time to try to refine, clarify and simplify current GD&T standards to fully empower such an “engine”.
3.    It is also time for more formal education in GD&T at mechanical engineering colleges and universities in order to raise awareness of its importance and enable more effective use.
Someday a truly smart “GD&T Encoding & Decoding Engine” will appear, and the world will change.  If our readers wish to comment, please let us hear from you.

Bill Tandler email:  bill@multimetrics.com
You can now view all QMT Magazine issues on your favourite tablet or smart phone.
Download the free Quality Manufacturing Today App from the Apple iTunes App Store or from QMT Magazine on Google Play.

Rob Tremain Photographer
Click above to see full page display and links to QMT articles.
Aberlink logo
Hexagon Logo 2
Nikon logo
Creaform logo
Bowers logo
MACH 2016 logo
Control logo
TCT Inspex 2016 logo