What do you think about this GD&T CBT training material?

Greetings:

I am wondering if anyone here has used this GD&T CBT (computer based
training courses) GD&T Trainer Professional Edition
Based on ASME Y14.5M-1994 from ETI http://www.etinews.com/trainer/. If
so, what do you think about it and is it any good for beginners? If
not, can someone please recommend a GD&T reference material for
beginners? I need to understand the why, how and where we use them.

Your additional comments and recommendation would be greatly
appreciated.

 

Haven't used it (don't need it), but the price is not out of line
considering what individual classroom training may cost. I've thought
about teaching a course myself for a set price per seminar rather than
per person, but I've got a problem with certification (which I think
costs a lot and isn't easy to get anyway). People would just have to
trust that I know my stuff ... and people tend to NOT trust any more
than absolutely necessary (understandably). I know just READING the
standard is a bit difficult, especially if it isn't written in one's
native language. It CAN be done (coming to understand the whys and
wherefors that way), but the standard isn't written as a tutorial. I'm
sure there are other written tutorials out there that are cheaper and
easy to read. Years ago Lowell Foster (member of the standard's
committee) used to publish his own interpretive book, but I don't know
if he published one on the '94 edition. It wasn't cheap, but it wasn't
$900 or anything close. I used to actually correspond with the guy
sometimes, but I lost track of him years ago and I don't even know if he
still lives (he wasn't young way back then). I can give you a few clues
about practical usage without getting too involved.

Some would argue that GD&T should be used in all possible situations. I
wouldn't say that, mostly because you're likely to run into problems
with machinists understanding the symbology. Instead I would say that
GD&T should be used mostly in VOLUME production documentation WHEN it
will buy you something in terms of:
1) clarity and/or
2) ensuring fit and/or
3) preventing unnecessary rejection of parts

Particularly locational tolerancing is useful for the 2nd and 3rd items
and it's the most widely used tolerance method of everything included
under the umbrella of GD&T. Using it, and especially granting bonus
tolerance for departure from MMC (in most conditions) or LMC (in a few
conditions like preventing hole breakout in bosses), is helpful not so
much to a machinist, but rather more to a QA department which has to
inspect (accept/reject) parts. It's helpful to a company subbing work
out when there needs to be a formal way to return parts if they don't
fit, based on a set of tolerances which give as much grace as possible
to the manufacturer without giving so much grace that ill-fitting or
non-fitting parts result. It's also useful in helping to create valid
statistical data for SPC (statistical process control) in which the
actual position of things like holes can be realistically correlated to
the probability of proper fit.

Contrasting locational tolerancing -- what was at one time referred to
as "true-position" tolerancing -- and normal plus and minus position
tolerancing (what I tend to refer to as Cartesian coordinate tolerancing
since it depends basically on an X tolerance and a Y tolerance):
Cartesian coordinate tolerances typically describe a square or
rectangular tolerance zone for the location of the feature (usually a
hole). The square or rectangle is as wide and high as twice the given
tolerances and is centered the theoretically correct location of the
feature (because you use +/- in front of the tolerance value). One has
to realize that the worst case for the position of the feature is the
hypotenuse of a right-triangle with adjacent sides that are equal to the
values of the X and the Y tolerances. In other words, by the
Pythagorean Theory if you have a +/-.005 tolerance in both X and Y axes,
then you have a possible actual worst case positional error of .00717
(the square root of X squared plus Y squared = the hypotenuse). If
you're actually counting on the feature being out only .005 worst case
then you may be in for an unpleasant surprise. One can say, though, if
parts will fit together when the hole is at the farthest X and also Y
tolerance allowance -- when it is actually .00717 from theoretical
correct position in the above example -- that you've wasted some
tolerance zone that could have been used and you may actually end up
having good parts (which would fit) rejected.
Positional tolerances use a round tolerance zone, which corresponds
with the real-world actual allowable zone in which features can be
positioned and create a fit scenario. It is more accurate because it
utilizes the entire usable tolerance zone instead of just part of it,
and/or it allows you no illusions that your parts may fit when they
actually may not. It also allows you to use a bonus tolerance for
conditions in which you've got the largest hole or the smallest pin or
stud or fastener (whatever) that goes through the hold. It doesn't take
much thinking to realize that when you've got a large hole it's going to
be easier to make mating parts fit together than when you've got a
smaller hole, simply because you don't have to be quite as precise in
the position of the hole. Using MMC you can say that if you get a
larger hole then you can allow the tolerance of the position to be a
little larger also. The opposite it true for LMC, which as I said is
useful when you want to prevent breakout. The smaller the hole in a
boss the less likely you'll get a breakout condition on the side of the
boss, right? So, if you give bonus tolerance for the smaller hole you
haven't increased the probability of breakout. Hopefully, needless to
say, giving a bonus tolerance for LMC might conflict with the purpose of
insuring fit if you're talking about a situation wherein the hole is
used to connect two parts together via screws or bolts or pins (or
whatever), which is why LMC is less often used than MMC for allowing
bonus tolerance. When giving LMC would conflict with the ideal of fit,
one shouldn't give any bonus tolerance at all. That is to say, the
tolerance should default to "regardless of feature size", which in the
standard that came out BEFORE the 1994 standard one actually had to
specify with a circle S (as opposed to a circle M for MMS or a circle L
for LMC).

Did that help?

Mark 'Sporky' Stapleton
Watermark Design, LLC
www.h2omarkdesign.com