November 15, 2015

The Black Arts (aka: GD&T)

I've heard people say Geometric Dimensioning and Tolerancing (GD&T) is a black art, and I agree it can almost seem that way sometimes.  My experiences over the past 20 years in the field of Coordinate Metrology demonstrate a widespread lack of understanding, confusion and even resentment for this engineering standard which was meant to make things crystal clear.  An international standard which will save everyone time and money. 

Before I delve into that, let me dazzle you with a brief history lesson on the origins of GD&T, as I understand them. 

Although there have been technical drawings dating back thousands of years, it wasn't until WWII, when the deficiencies of standard drawing tolerancing practices came to light.  It's said that Stanley Parker, of the Royal Torpedo Factory in Alexandria, Scotland , created a new positional tolerancing system which used cylindrical, rather than square, tolerance zones.  Now, before you Scots get all puffy chested...Parker was a Brit, as am I...well, half Brit.  I'm also half Irish, the half that doesn't care who invented it but thinks we should discuss over a pint.  Anyway, as I was saying, In 1944, the British published a set of drawing standards and went on to publish "Dimensional Analysis of Engineering Design" in 1948, which was the first comprehensive standard that used fundamental concepts of true position tolerancing.  

The Americans followed suit and developed their own standards, published in 1949, known as MIL-STD-8 and it's successor MIL-STD-8A, which authorized seven basic drawing symbols.  Eventually three different groups in the US who were publishing standards for drawings, the ASA (later becoming ANSI), the SAE and the Military, which led to much discussion and confusion over inconsistencies among the standards.  In 1957, the ASA (working with the British and the Canadians) published the first American standard for dimensioning and tolerancing, then in 1966, published by ANSI, Y14.5 was born.  This Y14.5 standard was first revised in 1973, where notes were replaced with symbols for all tolerancing.  Revised several times after that, we're left with what we have today, ANSI/ASME Y14.5-2009. 

So now that we're somewhat up to speed on the history of geometric tolerancing (or half asleep), let's get back to our original topic.  Why is GD&T so misunderstood? 

Now, even though I haven't come across any reference to the black arts, I'm not discounting it...but you can see it's been a source of turmoil and confusion for almost 75 years, so don't feel bad if you don't have a firm grasp on it just yet.  You're not alone. 

Since we've had a fairly consistant standard since 1973, you could conclude that it's simply a training issue, and that more people need comprehensive training in GD&T...and you wouldn't be wrong, but that's just one piece of the puzzle, in my opinion.  Yes, I believe companies should invest in formal training for all who work in the realm of design, manufacturing and quality, to bring about some basic understanding...and many companies do just that, but it's not enough.  It's not enough to sit in a classroom, or online, and follow powerpoint presentations, which leave you with a somewhat foggy understanding of the subject, with minimal retention. 

So you've got a general understanding of GD&T and what the symbols represent, but do you really know what that symbol means, what that number means, how it's calculated?  Do you know how they are applied in the real world?  Are you certain that's what best represents the fit and function of the part?  After all, the standard is meant to simplify matters and ensure things function as they were designed so long as they meet the criteria on the drawing.  If two mating parts both meet the drawing criteria, they should fit together right?  Not always.  To be certain parts will mate even in a worst case scenario, there needs to be a tolerance stackup analysis.  These analyses or simulations can be complicated and time consuming and my gut tells me that it's rarely done in the real world.  The designer's way around this, in many cases, is to apply much tighter tolerances than necessary...sort of a CYA approach.  I'm not trying to throw all you designers under the bus, just stating fact based on my experience.  If the designer uses very tight tolerances on both mating parts, they'll never have a fit issue...problem solved, right?  Not exactly.  The cost and time to manufacture the part just skyrocketed, which goes against the very purpose for the standard in the first place, which is meant to save us time and money.  Sure, the parts will fit together, but at what cost?  Old Stanley Parker would be rolling in his grave. 

Another pitfall we encounter due to a lack of understanding of GD&T is just plain incorrect application of the standard.  Too often, I have to work with engineering drawings which just don't make sense.  Sure, they look fancy and have lots of technical GD&T symbology, but looks can be deceiving and without always knowing the ultimate function of the part to be inspected, I'm left scratching my head trying to figure out what the engineer intended...because it sure isn't what's on the drawing.  If all else fails, I'll report some meaningful data (sometimes too much) about a specific feature in an attempt to compensate for unclear GD&T.  Again, lost time and money stemming from a lack of understanding.   

If you ask me, and you didn't but I'll tell you my thoughts anyway, the best way to a thorough understanding of GD&T is to work with it in the real world from the inspectors side.  It's not until you attempt to deconstruct the drawing and take real measurements, manually rather than with a CMM, based on the GD&T that you gain a clear understanding of what it all means.  I say "manually" because it forces you to make the calculations yourself rather than simply clicking an icon that will do the calculation for you.  The CMM will only cloud matters if you don't have some GD&T knowledge first, but that's a whole other topic.  Only then, will you have a clear understanding of how it all translates in reality.  I'm not aware of any GD&T courses that involve hands on inspection, and kudos if there are, but I believe some practical experience would greatly enhance the whole learning process and lead to much higher retention. 

A company I worked for early in my career would have all design engineers spend at least 2 weeks on the manufacturing floor.  They'd essentially be apprentices for that period, which allowed them to experience, first hand, the issues caused by poor engineering and/or drawings.  Not a bad idea huh? 

By now, you're either either knodding your head in agreement or preparing to spew venom in the comment box.  In either case, you can't deny there's an inherent problem with using a standard which is not fully understood by the vast majority.  My intention here is to not only highlight the issue, but also open up a dialogue on the topic.  Some of this is undeniable fact and some is just my opinion. 

Stay tuned for my upcoming blog series where I'll offer my take on some common GD&T challenges.

Leave a comment

Comments have to be approved before showing up.