Dear
ASTE:
It's the millennium! No planes fell out of the sky at midnight. No bank records disappeared and no computer took my house away. But I'm really depressed. I was up for the excitement of chaos.
And when it didn't happen, I stepped out onto the window sill of my third story bedroom and wondered if the laws of physics still applied. Could I fling myself out into the cool dark night, soar
over the treetops and glide into a parallel universe? Would I be sucked into the vacuum of the pitch black sky, embraced by the winds of time and caressed by a wreath of stars? Or would I simply
plummet to the ground and splatter like an over ripe melon, streaking the neighbor's lawn with my fluids? As I pondered the possibilities of the next thousand years, and my total lack of fear, I noticed
through the window opposite my own, that my wife was making out with the next door neighbor. I wonder how fast one would have to flap their arms to make it across the expanse of my fantasy to his.
Happy New Year!
Depressed in Decatur
Dear Depressed:
To you too.
Dear Mr. Meadows: I am interested in receiving training in GD&T toward certification. One of my fellow employees
recommended your courses. Will you be teaching GD&T classes (that are open to the public) anywhere in the Midwest in the year 2000? I have a copy of your book on GD&T with its workbook, but I would like to know when and where
you'll be teaching. Also, I would like to know what procedures and GD&T courses I need to get certification. Can you advise me in this regard? Thank you for your kind attention. K.J.H. Dear
K.J.H.:
I don't teach many open-to-the-public courses anymore. My entire calendar is booked for the year 2000 teaching and consulting on-site for companies and the military.
The only public course I am teaching is for the University of Wisconsin at Milwaukee on May 15-17. This course is entitled Functional Gaging and Geometric Tolerancing, and is almost entirely
applications based in which attendees learn a concept and then apply it.
The exam you wish to take to become certified is offered by ASME. It was written by the Y14.5.2
committee on Geometric Dimensioning and Tolerancing Certification of Professionals. It has almost nothing to do with applying geometric
tolerancing. It was written by reading a passage from the
ASME Y14.5M-1994 standard and then writing a question pertaining to the passage. If you want to pass the exam, my best advice is to read and memorize as much of the actual wording of this
standard as possible, then go in and take the exam. I had great success with this approach. The senior level exam is 100 questions and the technicians level exam is 125 questions. A greater
percentage of people have passed the senior exam. But, in fact, not a great percentage of people have passed either exam. I don't know what that means. I do know that had I not known the exact
wording of the Y14.5 standard, I would have missed many of the questions. You will need 85% to pass the exam. If you are having trouble with what the passages in the Y14.5 standard mean, then a course given by
a knowledgeable professional would help you. In fact most GD&T courses that are offered publicly are very simplistic, based on a ``this passage-and-or-symbol-means-this" approach. They offer few,
if any, applications, such as tolerancing assemblies, and are exactly the type of course that will help you pass the exam. It was nice hearing from you. I hope this helps.
Jim
Jim:
I understand that there are three gaging policies practiced throughout the world.
Policy 1 (U.S. Policy)
To accept most good parts, reject all bad parts, and reject a small number of borderline good parts. Policy 2
To accept all good parts, reject most bad parts, and accept a small number of borderline bad parts. Policy 3
To accept most good parts, reject most bad parts, accept a small number of borderline bad parts
and to reject a small number of borderline good parts. I have two questions:
1. In your experience, what percentage of companies in the United States use policy number 1 and what percentage use policy 2?
2. If a country chooses to use policy 3, what size gage pins do they buy
(plus or minus; or whatever they have available)?
A quick response would be GREATLY appreciated.
The books that you authored are a great source of reference for the GD&T community. In fact, I teach at a community college and put in a plug for your books in all the classes I teach.
Thanks,
D.R.K.
Dear D.R.K.: My experience going from company to company as a consultant tells me that there simply is no consistency in the tolerancing of gages, or knowledge that a recommended policy exists for ANSI
and another for ISO. In fact, gage houses and gage designers seem to be making it up as they go. Since tolerancing of anything, including gages, seems to be haphazardly done in all industries, I would
have to assume almost no one knows that the recommended policy for ANSI is Policy #1 and for ISO is Policy #2, and that they differ. Policy #3, I find, is most commonly not practiced with any thought to it at all. It is practiced by
people who first size their gage pins (either correctly or incorrectly) and then, as an afterthought, realize the gage pins need to be
toleranced, so they just assign a plus and minus tolerance to them.
Many make their own dedicated gage pins for specific situations.
But for those who do buy over-the-counter gage pins, the decision is often made right there, at the
counter, when the salesperson asks the question, ``Do you want plus or minus gage pins?" Thank you for the compliment on my books, and for any plugs you have given them.
You may be interested to know a gaging and fixturing standard is being finalized for geometric
tolerancing. It is B89.3.6. I am the chairman of the ANSI/ASME committee and will be submitting
it to the B89 (measurement) committee for vote soon. It is based on an older, much simpler standard called Inspection of
Workpieces-ANSI B4.4 published in 1981, but, recently retired, its contents to be absorbed into B89.3.6.
Hope this helps.
Jim
Follow-up response from
D.R.K: Jim: Thanks for the quick response. It's not the answer I was hoping to hear, but it doesn't surprise me.
We recently developed a Metrology program for this college (only the second Metrology program in the state of Michigan). In my opinion, I believe we need to make the students aware of all three
gaging policies, but the lab should include gage pins with a plus tolerance (per policy #1, the U.S. policy). We are teaching in the United States, we should be teaching according to the U.S. policy,
not the ISO.
Another instructor disagrees with me stating that we should purchase minus gage pins due to the fact that a company can't tell their customers their customer parts are bad because we 're
using plus pins. My response to that would be if we build a functional gage with minus pins, the gage is already outside of the specified dimensional limits (and prone to accept bad parts). As the gage
pins wear down, we are accepting more and more bad parts. I can foresee that if we accept a borderline bad part because we are using minus pins and the part fails in assembly, someone may
assume we need to tighten up the tolerances on the part being gaged, when the real problem is with the gage itself.
Thanks,
D.R.K . D.R.K.: For what it's worth, I agree with you. Jim Dear Jim:
We have several assembly drawings that specify alignment of two mating parts, a lid and a substrate (surface mounted), using a positional tolerance on the X and Y axes (from datums B and C). These
datum center planes are constructed by making the outside edges of each part datum features of size.
The question is, since the datum feature size dimensions are toleranced on the piece part drawing,
should they be shown on the assembly drawing as Basic, Toleranced or Reference? Using tolerances seems to indicate a dimension that needs to be inspected, when these parts have already
passed inspection at the piece part level. they are not `Theoretically Exact', so the definition of a Basic dimension doesn't seem to apply. And using a Reference dimension doesn't seem to be the
right thing to do either. What
do you think? Thanks in advance.
G.S. P.S. I've run about 50 people through GD&T classes based on your video series so far. It has really been helpful. Dear
G.S.
I would specify the dimension as a toleranced dimension, then if appropriate, add a note that says something like, ``Inspected at previous drawing level," or ``Inspected at piece part level." There is a
new rule in the Y14.5M-1994 standard that says the specifications only apply at the drawing level in which they are shown. This means you may wish to specify information at every level where it is
helpful. Having said that, specifying that dimension may be helpful as either a toleranced dimension, or as a toleranced reference dimension. As you said, calling a basic dimension is not appropriate.
There is a common misconception that not specifying something is better than a helpful specification because tolerancing something automatically means it must be inspected.
Actually, there is no such
rule that every toleranced dimension must be inspected. There is a B89 standard that should be coming out this year called Measurement Methods that will recommend writing a measurement plan
for every part that is designed that specifies how it will be measured, what it will be measured with and what will and will not be measured. You can get a jump on this by simply specifying a note for
this situation that says this dimension has already been inspected at the piece part level and need not be inspected for this assembly level specification.
The key to all these situations is to be absolutely clear as to what you want.
Hope this helps.
Jim
Jim: I have a question pertaining to datum planes derived from datum targets. The Y14.5 standard states
``since the orientation of the tolerance zone is established from contact of the part against a reference standard; the plane is established by the considered surfaces themselves." If a profile control uses a
datum plane established by datum target points on the same surface being controlled, does the datum plane establish the center of the tolerance zone, or just the orientation of the tolerance zone? Is the
profile zone unilateral, even though a phantom line is not shown, or is the tolerance zone equal bilateral, or is the tolerance zone simply a total wide zone without a starting point, that floats while oriented to the datum plane?
Help! I'm starting to lose sleep over this thing. My eyes are bloodshot. I mumble incoherently, and I hallucinate that I am being attacked by giant moths with rotten, bleeding fangs. My wife has taken
to sleeping in the spare room with the door locked. The other night I woke up in the kitchen with a can of whipped cream swirled over my chest, a butcher knife in one hand and my entire foot in a jar of sweet pickles.
What is the answer?!
P.H. in Dire Straits
Dear
P.H.: Lighten up! This question has been around since they were creating drawings on cave walls with
charred sticks. The question has several aspects, but the essence is always the same. First, any profile control (controlling size with basic dimensions from the datum) that references the datum
surface (forget about datum targets for the moment) as part of what is being profiled will lose half of the tolerance zone when it reaches the datum feature. The tolerance zone splits at the datum feature,
since the datum feature sets up the datum plane from which the zone is formed. This means that since the datum plane is formed by the highest points of the datum feature, then the datum feature
only gets half of the tolerance zone. There are two ways around this. he simplest is to exclude the datum feature from the profile tolerance zone. The other is to make the tolerance zone unilateral inside.
As to the specific question being asked, if we establish a datum plane for a coplanar profile control from datum targets that are on the surfaces being controlled (and let me say that I think this is not
usually necessary), then the tolerance zone is formed centered at the target points and splits there (unless it is designated as a unilateral zone with a phantom line). This means that the rest of the
surface (on which the target points exist) get the full tolerance, but the middle of the zone is formed by the datum target points and the plane they create.
If the targets are areas, I would suggest the datum plane not be referenced for the control of profile coplanarity that only relates the target areas to one another. Instead, just use profile coplanarity
without a datum and then reference the datum in other feature control frames.
That datum plane will be formed by the highest points of the areas simultaneously, just as though you set the part on riser
blocks (resting in each area) and the top of the riser blocks simulate the datum plane.
If the targets are areas and the targets are referenced in the coplanarity control for the datum target
areas (again, let me say that this is clumsy), then the datum plane still is formed by the highest points within each area. The coplanarity tolerance zone would then split at the datum plane, leaving only half
of the tolerance available for the target areas themselves (unless the tolerance zone is designated as unilateral inside with the appropriate phantom line).
The reason this is always argued over is that profile can use datums for either orientation or location or both. If a basic dimension is shown from the datum, then we know it is meant to control location
(and by default orientation). If a plus and minus tolerance is shown from the datum, then we know the datum is used only for orientation, not location. But with
coplanarity, since there is no dimension
at all given, just an understanding that the location dimension is zero, we are left to wonder if that dimension of zero is basic (giving the tolerance zone a definite location in space) or plus and minus
(giving us only parallelism). It has been discussed in countless ANSI meetings and always considered a basic dimension of zero (giving us location). But you can see it is an area that needs to
be put to rest with words in Y14.5 that address this situation specifically.
Hope this helps.
Your Pal,
Jim Dear American Society of Tolerancing Engineers:
I'm doing design work at a car component supplier (using a solid model CAD system). My education was in Industrial Design. In school, we engaged in assignments that were supposed to
foster and/or enhance creativity. If there exist, in fact, such activities, might there also be pursuits which will destroy whatever shred of creativity someone might still possess? If
so, obsessing over GD&T would certainly qualify! And don't get me started
about QS9000!!!
Sincerely Ticked Off,
Finn Dear Finn:
I have always said that design and tolerancing were two entirely different skills. I believe, as you seem to, that design is a purely creative endeavor and that tolerancing is a skill based almost entirely
in math and logic. A good tolerancing engineer must divorce himself or herself from all forms of emotion. It gets in the way. Tolerancing is not creative. It is deductive reasoning. The approach
should be, ``if this is true, and this is true, then I must choose these as datums and tolerance the assembly in this manner." For those who consider themselves sensitive, artist types, always
balancing on a tightrope, in danger of losing their creativity if exposed to the harsh realities of life, tolerancing could be the thing that sends you head over teakettle into the abyss. And worrying about
quality? No, no, not for the likes of you.
Better let the uncreative drone-types like me do the tolerancing and worry about quality. In your
world, everything is produced exactly as you perceived it. No need to clutter your creative juices with concerns over less than perfect parts, made by those of protruding brow and skinned knuckles.
Keep your head in the clouds where it belongs. Let us less creative, stoop shouldered australopithecines do the dirty work for which we were breed. We're just proud to serve. Just Kidding,
Jim |
