|
Re:
Notation vs. Notation of Specification
Jim,
I work with Pat (Ford GD&T Subject Matter Expert) and I would like your
opinion on a situation and some questions.
It seems that some folks view GD&T as notation only. My response
to this is that GD&T is more than notation; it is the notation of
specifications of form, fit, and their resulting function. If it was
just notation, GD&T would be nothing more than decoration. The
dimensioning and tolerancing on a drawing is the first indication of what's
required to verify that the part will fit and work in it's assembly or
system as intended. Manufacturing, tooling, gaging and assembly and
PPAP personnel depend on these specifications to form a complete and logical
system on each print.
I realize that it takes time to persuade skeptics and we've had our
successes. Pat and I facilitate your video series twice a year.
We try to consider our audience, what their specific needs are, how GD&T
can help them, but I'd love to have one or two more persuasion techniques in
the toolbox. From your experience, are there any
tools/techniques/tactics/strategies that can be put in play in these ongoing
discussions that can help designers and engineers to understand the
significance of these requirements? When you encounter a dismissive
attitude toward "that stuff you sprinkle on the drawing", what
have you found to be the most useful things that snap the discussion back
into focus?
No rush in answering. Any assistance that you can offer would be
greatly appreciated. Thanks.
Brian
FORD
NAT CAD Chassis
Brian:
I
appreciate your dilemma. It is frustrating to encounter the adamantly
ignorant and watch them lift their leg on all you hold sacred. You and
I know that these drawings are supposed to be legally defensible documents.
I get at least one call a month from law firms asking me to be an expert
witness against some company that hasn't defined their products well enough
to be understood by their vendors. The vendors want the parts they
have produced to be bought and believe they are in full compliance with the
specifications on the drawings, but the company says that isn't what they
meant at all. I always say no because, being a consultant, if you
testify against a company, they tend never to hire you again. Still,
it is interesting that in the industries I deal with, a person would
actually choose to be illiterate in the language we use to define the limits
within which our products must be produced to function.
The
truth is though, I don't get upset about it anymore. I only get hired
by companies that are committed to the use of good GD&T. If, in
the course of my work, I encounter a person who dismisses what I know or am
trying to teach them, so what? They are but a tiny stain on the fabric
of the corporate goals. I don't let them get in the way of the project
or training. Chances are, if they are that ill informed, they aren't
much good to the company anyway. They are a nuisance. Nothing
more.
I
don't view myself as a salesman of the topic. I just present it.
If they aren't smart enough to see the benefits, so be it. My vested
interest is always to teach everyone that is interested as much as I can in
the small amount of time I am given. As for the rest, well, I just
don't care about them. I am not an arguer. I am willing to tell
them what I know and if they either don't believe me or just don't want to
put it to use, I seriously doubt you or I will ever be able to change their
minds. Nor do I think it is worth the expenditure of time it would
take to do so, even if it was possible. I am rarely rude. I just
make it clear that they are entitled to their opinion, however misinformed
it may be. It is their problem, not mine or yours, unless you choose
to make it such. I honestly believe I would have burned out a long
time ago if I allowed myself to get caught up in the select few who choose
to dismiss what I have to share with them. So, if you want advice
(which I almost never offer), don't be offended by their poor judgment.
It probably has very little to do with you. It is, more than likely,
how they live their entire life. If they belittle what you know, in
their minds, they don't have to learn it. If they belittle everything
they don't know, they (in their own minds) never have to learn anything.
These precious few are not what the company can base their future on.
We hope they go away. And I believe they will over time, long before
GD&T ever does.
Good
luck.
Jim
Meadows
_______________________________________________________________________
Subject:
Measurement of Geometric Tolerances in Manufacturing
Howdy
Jim,
I think you have successfully converted me into one of your GD&T
groupies. After attending a few of your excellent seminars at the University
of Wisconsin, I have now spent the big bucks on your book "Measurements
of Geometric Tolerancing in Manufacturing". This book is a very
good read! Since I am one of your faithful disciples and desire to
achieve maximum value from this purchase I am hoping you will field a few
questions I have regarding the content of this book.
Question 1: On page 172 second paragragh you state "The RFS concept is
also useful for uniformity of fit, equal airspaces between mating features
and uniformity of pressure for hydraulics." Can you please
further explain using the RFS concept for "uniformity of pressure for
hydraulics."
Eagerly awaiting a response,
Colin
Colin:
I
apologize for the delay. I haven't been in the office much in the last
month and somehow your email got buried.
The
truth is that since RFS allows no bonus tolerance or pattern shift, it
creates more uniformity. In controls such as Position, since no
additional movement is allowed as a feature departs from its MMC or LMC
size, the features are better centered to the true position that both mating
shafts and holes share.
In
the case of a simple RFS position control on a clearance hole surrounding a
screw, the clearance hole can't move off to one side of the screw more as
the clearance hole grows in size.
In
the case of a pattern of holes surrounding, for example, a datum feature of
size such as a guide hole or guide pin referenced at RFS by the holes, the
pattern of holes may not shift off to one size of the datum axis more as the
datum feature departs from its virtual condition.
All
this means is, that since the pattern has not been allowed to shift off to
one side more, the inspector only buys parts that can be assembled on
center. If the datum feature of size had been referenced at MMC, the
inspector would have been able to buy patterns of holes that had shifted off
to one side of the datum axis. To fit these kinds of parts into an
assembly, the assembler would have to push the weight of the part back in
the opposite direction (than the pattern shifted) in order to get the
individual parts to fit together. This would create an inequality in
the airspace of one part fitting inside of the other or in the balance of
the assembly. This is especially bad for parts that need uniformity of
air for pressure, such as seals, since the space on one side of the seal
would be greater than on the other side, and it's also bad for parts that
have to spin and balance.
If,
for example, a seal had to expand as heat was applied to the assembly to
fill an airspace, it may not be able to expand quickly enough in a given
period of time to fill the unequal airspace (allowed by datum features
referenced at MMC) and cause the product to fail. Failures such as
this could be leaks or even explosions (picture the space shuttle).
Spinning parts forced to be assembled off to one side may spin unbalanced.
With sufficient speed and off to one side enough, the balance could be so
bad as to cause this assembly to fail (picture the turbine blades on a jet
engine breaking apart and entering the cabin of a commercial airliner).
There are many more examples, such as ill-fitting automobile panels with big
gaps on one side and no gaps on the opposite side, but if you need to
discuss this more, call me sometime when I am due to be in the office.
The person answering the phone here can tell you when that is.
Hope
this helps,
Jim
_______________________________________________________________________
Subject:
Measurement of Geometric Tolerances in Manufacturing Textbook Question
Jim:
I
am looking for a manual that provides explanations and illustrations on the
use and understanding of GD&T as it relates to the setup and
measurements using coordinate measuring machines (CMM) and/or measurement
setups. Your manual titled "Measurement of Geometric Tolerances in
Manufacturing" may be what I am looking for. Am I correct in this
assumption? I have attended your classes in the early 90's and
have the workbooks from those classes. I have found them useful over
the years.
Mike
Metrology
Engineer-Hewlett Packard
Dear
Mike:
My
textbook on Measurement of Geometric Tolerances in Manufacturing provides a
measurement look at geometric tolerances. It shows what the language
means and how to measure geometric tolerances using many different
measurement tools. It includes information about proper datum
establishment, and measurement of geometric tolerances using coordinate
measuring machines, but does not focus solely on CMM's. It shows how
to design, dimension and tolerance gages and fixtures. It shows the
proper set-ups and measurement procedures for surface plate measurement,
optical comparators, roundness checking machines, how to measure runout and
concentricity and a wide variety of other geometric tolerances in a wide
variety of ways. It shows how to write a measurement plan and what
some of the measurement uncertainties may be. It has a lot of case histories
related as stories about measurement and quality problems encountered by
other companies that I have consulted for over the years, and how they dealt
with them, both good and bad.
So,
in short, if what you are looking for is a book that deals with CMM
measurement as its main focus, this is not the book for you. But, if
you are looking for a book that deals with CMM measurement as a part of a
broad spectrum of measurement techniques to teach what geometric tolerances
are and how to find out if parts are in or out of tolerance, a book that is,
in spots, actually fun to read, this is the book for you. This book is
450 pages long. This book is sold by the publisher Marcel Dekker,
Inc., but currently www.barnesandnoble.com
has a great price (almost $100 less than the normal price) for this book.
I don't know why, other than they must have bought in bulk.
I
have another textbook that is longer (624 pages) and can be purchased with
an accompanying workbook (another 340 pages) that is a more comprehensive
look at GD&T in all its aspects. The textbook is called Geometric
Dimensioning and Tolerancing-Applications and Techniques for Use in Design,
Manufacturing and Inspection. This book is more technical and broad in
its approach touching on all areas of GD&T except tolerance stack-up
analysis. These books can be purchased from Marcel Dekker,
www.amazon.com
or www.barnesandnoble.com.
I
also have a new book, Tolerance Stack-Up Analysis, (350 pages) that covers
this topic for stack-up in assemblies, but even has a chapter on single part
analysis. I believe it to be the most comprehensive book on this topic
on the market. This book can be purchase directly from us by calling
(615) 824-8644. You can also learn more about all
these books at our website at www.geotolmeadows.com.
I
hope this helps.
James
D. Meadows
_______________________________________________________________________
Hello
Jim:
In
going through notes from your class at my company around 1992, I have the
following question.
When I have a part tolerance +/- .005, is it necessary to specify a flatness
of .005. Flatness is a refinement of the size and recommended is half or
more of size tolerance, since size tolerance is shared by both sides. Also
for the same part, assuming .005 flatness is specified, is it necessary to
specify a parallelism of .005,as that is what is left?
Thank
you very much for your time and help.
Sundar
Applied
Materials
Dear
Sundar,
It's
a judgment call on your part as to whether or not to refine the control your
size limits give you. A rigid feature of size with a tolerance of +
.005 is allowed to be out of flat, straight or parallel the entire size
tolerance of .010. In fact, if you visualize a flat plate that
uniformly curves similar to a banana in shape, it could be out of flatness
the entire .010 on both sides of the part at once. There are only two
requirements to meet the size limits. One is that the part fit a GO
GAGE made at the MMC size, and the other is that it measure locally (like
with a micrometer) not in violation of the LMC. So, in extreme cases,
the size tolerance only controls flatness of any one surface to within the
full amount of the size tolerance (not necessarily shared between the two
sides of the part). In concept, any refinement of flatness or
parallelism controlled by the size tolerance that is smaller than, in the
case you describe, .010 is a refinement.
What
you do to refine the flatness and parallelism, if anything, is up to how
much control you think you need on the part.
Hope
this helps.
Jim
Meadows
_______________________________________________________________________
Subject:
GD &T Question
Jim,
a
question if I may......?
I
have attached drawing 8522-10. My question is on the 2X 1.525 +.005
-.000 true position to .005 at MMC to ABC. I am taking it that the 2X
.927bsc dimension is my X measurement and the 2.706 bsc dimension is my Y
measurement that need to be held to the .005 MMC. My question is since
the .927bsc dimension is coming from a hole can I add any bonus tolerancing
for the holes deviation from MMC or do I strictly use the 1.525 cut out
feature to control my bonus tolerancing? I am thinking I have to use
the cutout's MMC only.
Thanks!
MaryAnn
TriContinent
Dear
MaryAnn,
The
basic dimensions are just a way of showing the relationship between all the
features being positioned and the datums they are positioned to. Basic
dimensions accumulate no error, unless you have more than one set of datums
on the drawing and switch in mid-stream. So, it really doesn't matter
what the basic dimensions come most immediately from, as long as you can add
them together or subtract them from one another to let you know their
distances from the datums they are positioned to. In short, that means you
are right, the only bonus tolerance you get is from the feature under
consideration. Each feature has a right to only its own bonus
tolerance gained as it departs from its MMC.
Hope
this helps.
Jim
_______________________________________________________________________
James,
Good morning, I work in Supplier Quality ---. I have a question I was
wondering if you could help me answer.
With
a unilateral True Position. (only measured in X axis), why is the deviation
doubled? Is this call out calculated and seen to be basically like a
profile?
Thank
You.
Scotty
Bosch
Dear
Scotty,
The
position tolerance might not apply in both the X and Y directions, but the
tolerance is still equally split on either side of the basic dimension (or
true position), half the tolerance existing on each side. So, the
deviation from true position, if it is to be compared directly to the
position tolerance that appears in the feature control frame, must be
multiplied by 2.
Hope
this helps,
Jim
Meadows
_______________________________________________________________________
Subject:
Pins for locating assemblies
Jim,
I am a mechanical engineer at Varian Medical Systems in Palo Alto Ca and
have attended several of your on site classes as well as viewing your tape
series. I have a question regarding the use of pins to locate
precision sub-assemblies within our medical machine. These assemblies
must be removable for maintenance a few times during the life of a machines
(approximately 15 years). For all intensive purposes, assume
that I would like to accurately assembly one plate relative to the other
(plate "1" to plate "2").
The
solution I have proposed is to use a locating pin and a diamond pin on plate
"1" and 2 holes on plate "2". I have assigned the
flat mating surface as the primary datum for both parts. Because
the fit must me accurate, yet removable in the field I have suggested a
"locational clearance" fit, based on American National standard,
for the secondary datum's of plate "1" and "2". I
have assumed that the tolerance method for both pin and diamond pin is the
fixed fastener formulas i.e. (MMC hole -MMC shaft) = Geo. divided
between mating features. The secondary datum for part "1"
and part "2" would be the pin and hole respectively, with a
perpendicularity control to the primary datum and tolerance at MMC according
to the fixed fastener formula. The tertiary datum for part
"1" and part "2" would be the diamond pin and hole
respectively, with position control to the primary datum, secondary datum at
MMC, with a tolerance at MMC according to the fixed fastener formula.
After
I calculated the tolerance for the position control for the tertiary datums
of part "1" and part "2", I found that the tolerance is
unrealistic, even with a bonus tolerance. Therefore, I decided to make
the preferred metric fit, for tertiary datums, larger such as a "close
running" fit according to the American National Standard and I still
used the same fixed fastener approach to the tolerancing. In terms of
accuracy of the assembly, this larger fit should only sacrifice a slight
angle offset at assembly.
My
question to you is this...
1.
Do you agree with this approach?
2.
What additional recommendations would you make to this approach?
3.
Am I correct in assuming that the gd&t of a diamond pin should be made
with the fixed fastener formula?
4. Does the diamond pin geometry give you any additional bonus in
position tolerance that is not captured in my approach, or does the diamond
pin geometry simply avoid redundant constraining?
I would greatly appreciate your response to this matter as your insight is
valuable to our organization. Thanks.
Chris
Dear
Chris,
1.
Yes, Chris, I think this is the same type of approach I would have used.
2.
I have no additional recommendations at this stage of the design. Of
course, now you have to reference the datum reference frame you have
created. I assume there will be positional controls that will follow
and reference these datums. To recommend additional controls, I would
have to know more about the parts. But it seems you have it well in
hand so far. I have no reason to believe you won't continue to do
well.
There
is one comment, though. If you were to look in my textbook (the yellow
one) on page 150, you would see that I follow the same type of approach you
are using, but instead of a diamond pin on the mating part, I use an
elongated hole on the part with the holes on it. This approach is
widespread in the automobile industry and would work equally well as the
approach you have chosen. It isn't necessarily better, just more
common (well maybe a little bit better). The diamond pin approach is
actually more common to gage design than mating part design.
3.
Yes, these are all fixed fastener assembly conditions.
4.
As I mentioned in number 2, the diamond pin approach is actually more common
to gages than mating parts. But, no to the additional bonus tolerance.
I believe your goal here is to treat the pin as though it is cylindrical and
to just take advantage of the reduced section of the pin to make certain it
isn't controlling location, but instead just controls rotation (just as an
elongated hole on the mating part would accomplish, if used instead of a
diamond pin on this part-Hint-Hint).
At
any rate, it appears you are doing a great job. You must have had a
good teacher.
Jim
_______________________________________________________________________
Jim
Hello from Austin Texas.
Paul
from 3M Co. I had a class of yours many years ago here. Anyway,
I have a question about Composite Tolerancing. In the Top Tier, my
pattern is related to the 3 Datums and I have no problem inspecting my part
to that. In the bottom tier, I don't know where to start on my pattern
to measure their position. I have a pattern of 400 holes, 20 x 20.
Do I start(zero) in the corner, in the middle, ?? What do I skew
on? Is there a correct procedure as I cannot find it in any of the
Standards.
I am using a vision inspection system which will skew and zero on whatever
feature(s) I tell it to.
Thanks for your help.
Paul
Paul,
There
is no rule on where to start or what to clock to when inspecting the feature
relating tolerance zone framework (lower level-hole to hole tolerance).
The
truth is that the data is often collected as to the location of the holes to
the datum reference frame in the pattern locating tolerance zone framework
(the upper level-datum to pattern tolerance), then the data is released from
the datum reference frame and manipulated to see if it falls within the
tighter lower level feature relating tolerance zone framework.
If
you want to reset the part when checking the lower level control, you can.
And you can start anywhere, then just rotate and translate the collected
variables data until it checks good, or you find out it is a bad part.
Still,
you might consider just setting the part once (to the datums), collecting
the variables data, see if you meet the upper level tolerance while related
to the datums, then release the data from the datums and see if you can
manipulate (translate and rotate) the data to meet the lower level (tighter)
hole to hole tolerance.
Jim
_______________________________________________________________________
SUBJECT:
DATUM QUESTION
Mr.
Meadows,
How
are you? I hope that you are doing well. I am sure that you are
a very busy person but I would like to ask you a question. I have your
textbook as well as the Y14.5M standard and unfortunately neither resource
addresses the issue at hand.
On
page 110 you show figure 6-33 illustrating datum targeting for Datum A (as
well as the others). Let's say that the surface of datum A is, in
actual use, also a sealing surface and thus must be controlled with a
flatness tolerance. What is the best way to show the feature control
frame with the flatness tolerance and to show that it is for the entire
surface, not just the datum target areas? On an edge view of this
datum surface, would the datum feature symbol still be shown (figure 3-2,
page 42, Y14.5M)? I read through pages 83 and 340 to 343 but those
pages always show discontinuous coplanar surfaces, not a continuous surface
with datum target areas.
Many
thanks in advance.
Marcus
Siemens
VDO Automotive
Marcus,
An
edge view would be the best place to show the flatness control. The
fact that the surface you are making flat also has datum targets on it are
separate issues. Just show them as separate things. When datum
targets are used, there is no need to show the datum feature symbol at all.
But if you want to show it as a clarifying redundancy, show it separate from
the flatness control. Otherwise there might be some confusion, since
the entire surface must be flat, but only the datum targets are to be used
to construct the datum plane.
Now,
I have a question for you to ponder. If the entire surface is a
sealing surface and therefore must be flat, why are you assigning datum
target areas to it to construct your datum plane from. Normally, in
these situations, it would be most functional to use the entire surface as a
datum feature, not datum target areas. There is no need to respond to
me with the answer. It is just to get you to think about the fact that
sometimes we get talked into doing things like this by manufacturing or
inspection to mimic the way the part is set up by them. But, if that
isn't how the part functions, by assigning datum targets, we are sanctioning
them to measure the part in a non-functional way. And, somehow, in
this non-functional (but easy) measurement, they are supposed to determine
whether or not the part will function.
Nice
to hear from you. Keep in touch.
Jim
Meadows
_______________________________________________________________________
Subject:
GD&T Question
Dear
James,
I
attended your class last September at Smith and Nephew, and I am finally
applying GD&T extensively to some highly contoured plates that our
development team is working on. In section 6.2 in page 93 in your
book, you have written "It is now legal to identify a compound curve or
contoured surface as a datum feature." Would you be able to
provide me with more detail on how to technically identify this feature on a
print? If you could point me in the direction in the ASME Y14.5M-1994
Standard, I would greatly appreciate this.
I
believe you will be at Smith & Nephew in the next week or so teaching a
class so I could meet with you at this time to discuss this issue.
Your help is greatly appreciated. Please give me a call at your
earliest convenience.
Darin
Smith
& Nephew, Inc.
Darin,
On
page 68 of the Y14.5M-1994 standard, section 4.5.10.1 Mathematically Defined
Surface, it says, "It is sometimes necessary to identify a compound
curve or a contoured surface as a datum feature. Such a feature can be used
as a datum feature only when it can be mathematically defined and can be
related to a three-plane datum reference frame. In such cases, the
theoretically true geometric counterpart of the shape is used to establish
the datum." Unfortunately, Y14.5 doesn't show an illustration of
what this would look like on a design drawing. However, there is a
brand new ASME standard, ASME Y14.43-2003 Dimensioning and Tolerancing
Principles for Gages and Fixtures, which shows an illustration of this
concept. This standard is available to the public from ASME starting
this week. I happen to be the chairman of this standard and, as such,
received a copy of the published standard a couple of weeks ago. I am
looking at page 80, figure B15 Irregular Closed Feature Used as a Datum
Feature. It shows an irregularly shaped hole on a plate that is used
as a datum feature. It is defined with profile of a surface and, since
it is an oddly configured hole, it is positioned using the positional
BOUNDARY concept, then it is identified as a datum feature. Other
features are then positioned to it. It then goes on to show how this
hole would be represented at its virtual condition for a gaging procedure of
the part. I would be happy to show you this illustration when I get there
next week.
I
also have a much simpler (and less complete) illustration of a curved
surface (not a hole or a shaft, just a curved surface) used as a datum
feature in my yellow workbook (the one that accompanies the yellow hard
cover textbook you referred to). I believe it is page 6-11 in that
book. It might require a little explanation, though.
I'll
look forward to seeing you sometime next week.
Jim
Meadows
_______________________________________________________________________
RE:
Profile Tolerancing
James:
It's
been a long time since I was at your CTC training back in 1998. How
are things ?
I have a quick question for you. When the CMM verifies profile of a
line, how does it calculate the result ? Is it true that the CMM takes
the range value and uses that as your true profile. OR do you take the
worst value and multiply it by two ?
Example:
Spec: Profile 1.0 mm
Data: 0.5, 0.6, 0.4, -0.3, -0.5
Results: The CMM takes the highest and lowest and calculates the
range...answer = (0.6 + abs(-0.5) = 1.1
Or do you take the highest number (worst case) = 0.6 and multiply by 2 = 1.2
The answers are different.
Can you please help me...
Thanks,
Paolo
Quality
Engineer, Siemens VDO Automotive
Paolo,
CMM
software has no standard way of calculating tolerance compliance.
Every software program has different algorithms. Software companies
don't share their algorithms with each other. They believe their
algorithms are what makes them unique and gives them an advantage in the
marketplace. My advice is to call the company that programmed your
software and ask them.
Jim
Meadows
Jim
I
have already contacted the CMM company on this. I was wondering based
on your expertise is my thought process correct or is the profile tolerance
calculated differently.
Can
you elaborate on the mechanics of the calculation.
Thanks,
Paolo
Paolo,
The
CMM approaches on this are either right or wrong depending on the profile
control. If the surface being controlled is located by basic
dimensions using an equal bilateral tolerance, then taking the greatest
deviation from the basic dimensions and multiplying it by 2 is correct and
it would be compared to what is in the Profile feature control frame.
If it is greater the part is bad. Equal to or less than the profile
tolerance and the part is good.
If
the profile tolerance is equal bilateral, but there is no location by basic
dimensions involved and instead it is just to control the shape of the
surface, then the approach where the range of deviations is determined could
be correct, if the surface is just planar in shape. But if the profile
is of an odd configuration, you may be tracking the deviation from certain
radii. In this case, since the tolerance zone may be moved about to
contain the surface, having a huge plus deviation and a small minus
deviation may have the effect of violating the profile zone anyway, even if
the difference between the plus deviation and the minus deviation is not
greater than the profile tolerance. Since you are trying to calculate
the deviation from the basic radii, a more sophisticated approach may be
needed, wherein you graph out the deviations (or simulate this in the CMM
software) to try to manipulate the surface into the tolerance zone.
I
think this problem is more complex than can be explained in the two
approaches for the CMM that you described. Certainly, it is much
simpler if the Profile is meant to control not only form, but also size,
angles and location. In this instance there is no manipulation of the
data. Once the datums are aligned, the probed points on the surface
profile are either in the tolerance zone or they are not. But even
this falls apart if there is a datum feature used with a maximum material
condition symbol next to it in the profile control. This would allow
shifting of the data to within the allowed deviation of the datum feature of
size from either its virtual condition or its maximum material condition, as
appropriate. The chance of the CMM program handling this shift of data
to within the allowed amounts is not good.
I
know this is probably more than you wanted to think about. And I
haven't even mentioned the situation where the tolerance is unequal
bilateral or unilateral.
Generally,
CMM's try to simulate close to what the geometric concept asks for.
But as you can see, it requires a great deal of knowledge on the part of the
programmer to even know all of the possibilities of what can happen and what
to look for. So, my answer is to ask the CMM programmer what he did,
and then you will have to determine on a case by case basis whether or not
this is the correct approach. The question with CMM programs isn't if
it is right or wrong. The question is, "How wrong is it?".
And is this a level of correctness that you are willing to live with.
Or maybe ask yourself, "In this situation, I need the program to do
this. Is it doing that?" And, "If not, is it good
enough?".
Now,
I know you must be frustrated. Sorry. I tried to spare you this
with my first answer. The world of CMM algorithms is a real can of
worms.
Jim
Thanks,
James, for your input...We can now view this a little differently.
Paolo
_______________________________________________________________________
Re:
question for an expert
Hello
Jim,
Having
been one of your past students (along with a co-student by the name of
Patty), while an employee of Ford Motor Company, I've had the pleasure, and
luxury, of referring to your text in "GEOMETRIC DIMENSIONING AND
TOLERANCING - APPLICATIONS AND TECHNIQUES FOR USE IN DESIGN, MANUFACTURING,
AND INSPECTION" on numerous occasions.
Being
that our manufacturing facilities are global, numerous geographical
locations are in need of the expertise contained within your book.
I
would appreciate your permission to scan (copy), and send subject matter
(i.e. Section 6.5, FLEXIBLE PARTS) to a limited number (2) of our
manufacturing co-workers, with a footnote giving credit to your book.
With
my highest level of respect,
Michael
Visteon
Design Analyst - CAD
Michael,
It's
good to hear from you. I appreciate the compliment. I don't
seriously think anyone will mind if you scan the section you mention, credit
where it comes from, and send it on to a couple of co-workers. But,
the truth is, I don't own the rights to that material. It is owned by
the publisher, Marcel Dekker, Inc. So, although I can't give you
official permission to do it, it seems like your act could only generate
more interest in book sales. Keep in touch.
Jim
_______________________________________________________________________
Subject:
question
Mr.
Meadows,
I am writing this e-mail to you to ask an expert in GD&T a simple
question. Can MMC be applied in a positional tolerance to a threaded hole.
and if so, how?
I believe that there is no logical way for MMC to be applied to a threaded
hole since the logic is that a bigger hole can be "out of
position" more than a smaller hole and still achieve fastener fit.
Making a threaded hole "bigger?" (if there is such a thing) would
not allow any "forgiveness" to
the position of the hole since the threaded feature will center the hardware
anyway.
I know you are not in the business of giving away "free"
information just for kicks, but I would like to be able to site an authority
on the subject of GD&T before I take a stand on this specific practice
and I thought that you might empathize with me (and my frustration) on
subjects like this.
Thanks for any help you can offer me on this subject.
Sincerely,
Guy
Guy:
Every
example in the Y14.5 (Dimensioning and Tolerancing) standard and in the
Y14.43 standard (Dimensioning and Tolerancing of Functional Gages) that
shows threaded holes uses the maximum material condition symbol. I am
the chairman of Y14.43 and in a joint committee meeting recently, one of the
members of another standards committee asked the same question you are
asking. Members of both Y14.5 and Y14.43 leaped at the chance to
explain why we do this. The answer is this: As the pitch
diameter of the threaded hole grows, an airspace is created between the
screw and the hole during assembly. This allows the screw being
inserted to wobble inside of the threaded hole an amount that is not easily
quantifiable, but exists nonetheless. An inspector engaged in
measuring the positional tolerance of that hole has a couple of choices.
One is to use an attribute gage known as a functional gage. In order
for this gage to be allowed to use a fixed size gage pin which duplicates
what the assembly uses (a fixed size screw) a maximum material condition
symbol must be used in the position of the threaded hole. If the
threaded hole is positioned at RFS the gage would be forced into using an
expanding gage pin which is not considered feasible and is also not
reflective of functionality (since the screws used in the assembly don't
expand). The fixed size gage screw utilizes any airspace created, as
the pitch cylinder of the hole grows, to accept more parts. Remember,
the goal of the gage is to determine whether or not the part will assemble,
so the air/slop created acts much like a class of fit on threaded holes to
allow a less restrictive, but entirely functional, part to pass.
Now,
if the inspector wants variables data, he may choose to use CMM's and
probes, or even dial indicators to probe the gage pin as it projects from
the hole (hopefully the threaded hole uses a projected tolerance zone as
most should). In this case, using a fixed size gage pin is probably
not best because the airspace created by the growth of the threaded hole's
pitch diameter would allow the gage screw to wobble in the hole while the
inspector is trying to probe it. So, he may choose to ignore the fact
that there is a maximum material condition symbol in the position control
and use an expanding gage screw instead. This will create a more
stable measurement, but he loses the additional tolerance that he was
entitled to. Since this additional tolerance was negligible anyway, it
is no great loss.
As
long as the MMC symbol is used on the threaded hole, the inspector has the
chance to take advantage of this small amount of additional functional
tolerance if he decides to use a functional gage. Even if he chooses
not to take advantage of this extra tolerance today, the drawing is
correct/functional and allows him to change his mind in the future, should
he decide to do so. The goal is to never force an inspector to inspect
parts in ways they do not function (in this case with expanding screws) and
consequently to reject a fully functional part.
I
hope this helps. I know convincing people to allow a fully functional,
but not easily quantifiable tolerance is a hard sell, if they already have
their minds made up not to. Still, the mindset is the important thing
here. Always tell the truth on drawings. If the part works in a
certain way, try to allow the inspector to inspect it that way.
James
D. Meadows |
