ASME Y14.5m application and general drafting stds 3

[DesignBiz]

1. The attached tube dwg is the current template for tube dwgs. I find it obviously in error in many respects, in particularly the use of A-B as a single datum. (Ref ASME Y14.5m 1994 4.5.7.1 and 4.5.7.2) In this case the ends of the tube are A and B and these features of size are at compound angles to each other. The in-house consultant rationalizes this callout somehow in his thinking and claims it is easily understood according to ASME Y14.5m 1994. I couldn't disagree more.

2. The profile callout via default note completely defys ASME Y14.5m 1994 in my estimation. (ref ASME Y14.5m 1994 6.5.1 paragraph (a) regarding required view or section for a profile callout.

3. Basic drafting standards seem to be a foreign concept.

Comments invited.

D-Biz
Sr. Designer (Auto-Aero Mechanical)
NX4 / Team Center user
25+ yrs experience

 

[DesignBiz]

Thank you MechNorth and Ringster for your inputs !

You were able to view the 2 JPEGs I posted?

This is actually what I had hoped for from this website

People who are interested in ASME Y14.5m and can explain their position using the standard and geometric
construction. It is interesting that so much of the standard can be interpreted differently by knowledgeable people.
This was brought to my attention approx. 1996 when at Delphi in GDT training for the 1994 standard, the certified
GDT instructor mentioned that he and "his buddy" could debate some things aspects for hours. Boy, that blew my
naive thinking regarding when I had first been "given instructions", for 1973 stand by Don Buckman at GE Evendale,
that this was a common language for all to understand drawings the same way.
It looks like MechNorth that you are a certified instructor also. A future goal for myself.

At this point in my career I have been trying to go to the current standard and it's definitions for as simple interpretations
as possible. My problem still persists even though MechNorth does give a logical explanation as well as "food for thought"
regarding the incompleteness of the standard and the need for advanced callouts.
Axym originally makes this claim, however, this kinda gets me excited when leaving me
with a comment that the standard is lacking without suggesting a remedy. Ok, let's agree that it does not encompass all possibilities. Now what? Anyone can
apply their advanced callout without explaining it in an amendment? Might as well go back to the old Cartesian plus/minus system
at that point or else we have arrived at the tower of Babel! No offense meant to anyone. Actually I find the standard and people's interpretations
fascinating and "thought provoking regarding designs". I'll keep it with its flaws.

Now back to my persisting problem regarding the A-B callout on
the drawing I posted. I look at the datum definition; plane, axis or exact point. I see that a datum reference callout in a Feature Control Frame
is for a "single" datum. I see that I cannot geometrically create a single datum for two features of size at compound angles as in the JPEG I posted.
From the same JPEG I believe it can be seen that a DRF cannot be determined from those two cylinders which axis are at compound angels.
Can't find the DRF origin. At this point the callout cannot meet the standards definition, unless there is an amendment to our standard here which explains
as MechNorth endeavors to do from his point of view.

My position is that advanced type callouts can only be used when amended to explain what is lacking in the standard for that callout
and/or deviated from.

FYI, my 9th grade Biology teacher informed us that "ex or X" is the unknown, and "spert or spurt" is a drip under pressure, thus;
Expert = an unknown drip under pressure. Just kidding !

It seems like a conic for all practical purposes should and could be added to the definition of feature of size.

I think I have concluded that our in-house "consultant" could be correct regarding the A-B callout if we would explain it in an amendment.
He doesn’t want to do this though.... he fears the dreaded "committee" to get it into our company standards.

Any one up to a profile callout as a "general note". We have it, and guess what? I disagree.


DesignBiz

"Quality is in the details"

 

[axym]

Wow, interesting thread. I have some comments about the original post’s issue of the A-B datum feature and will comment on some of the other issues later.

I agree with MechNorth that “multiple datum feature” is the correct term, not “compound datum”.

The functional situation that MechNorth described makes sense. The design intent might be to hold the part in a fixture that contacts datum features A and B without giving one of them precedence over the other. In other words, the degrees of freedom are simultaneously constrained by the two datum features as a group. It’s the equivalent of the scenario in Fig. 4-21 of the standard, except with datum features that are not coaxial. The key thing is that neither datum feature lines up exactly with its simulator (neither one constrains particular degrees of freedom on its own) . The multiple datum feature approach is useful (and necessary) in this situation, because the desired constraint wouldn’t be possible if one of the datum features was referenced as primary. For example, if datum feature A was referenced as primary, it would be obligated to constrain 4 degrees of freedom. Very different. Some of the posts favored the “conventional” primary-secondary-tertiary sequence over the multiple datum feature approach, just because the features are skewed. I don’t agree with this. I agree (again) with MechNorth that the intended function should dictate how the datum features are configured. I’m no expert on tubes, but it would seem that tubes of this type would often be aligned at both ends simultaneously and not aligned to one end first.

So is it legal to reference a multiple datum feature comprised of non-coaxial cylinders? This is one example connected to a more fundamental question. There seem to be two schools of thought – one says “it’s not legal because the standard doesn’t specifically mention it or show an example of it” and the other says “it’s legal because it’s a logical extension of a Y14.5 principle, there’s just not a specific example shown”. In this case, I would say that it’s a valid extension of principle. Paragraph 4.5.8 reads “multiple features of size, such as a pattern of holes at MMC, may be used as a group to establish a datum when part function dictates”. It then goes on to describe and illustrate one example with a square pattern of 4 parallel holes, referenced as a secondary datum feature at MMC.

I would say that the two non-coaxial holes are another example of multiple features of size and that we may use them as a group if part function dictates. In the situation described above, part function dictates.


Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[DesignBiz]

Axym,

I can agree with the "logical extension" as you put it, when the logic is relative to the part function and how it
should be restrained in some cases.

In the case of a hole pattern as a referenced datum, I can determine the center of that pattern geometrically
and create a DRF that fits the definition in section 4 of the standard. In the case of cylindrical features of size at compound
angles to each other I cannot, unless someone explains it either like yourself or MechNorth does. I would wager that
if you showed this callout to your average technically proficient GD&T person that you would not get most to understand
it based on the definitions of the standard without further explanation.

My only point is that if you do something like this, it merits an amendment to provide that explanation. Not necessarily in
the standard itself, rather an amendment by the company who owns the drawing. After all the whole idea here is for everyone
involved to understand and inspect the part in the same light.

If you cannot agree with my thinking then I suppose we can just agree that we disagree.

Thanks to everyone for their input, I would like to start a new thread on "profile".


DesignBiz

"Quality is in the details"

 

[MechNorth]

Ringster, no flaw in the reasoning, just shortcomings in the standard. In the absence of clearly defined orientations for the first two mutually perpendicular planar datums (intersecting at the datum axis), the designer arbitrarily sets the orientation. The designer also then arbitrarily sets the orientation of the third datum plane, mutually perpendicular to the first. These "arbitrary" orientations need to be documented on the drawing, and manifested repeatably on the simulator(s).

Again, these are extensions (distant extensions, but extensions none the less) of first principles. DesignBiz, don't worry about documenting this in an amendment; it belongs on the drawing itself. 1.1.4 Figures: ... The absence of a figure illustrating the desired application is neither reason to assume inapplicability, nor basis for drawing rejection. ...". The best place to communicate design (in this case, the location and orientation of datum planes) is on the drawing, not in auxiliary documentation. (I'll sketch something up and post it soon).

I was able to view the second graphic; the first didn't even list on posting (for me at least).

Evan, good reference to 4.5.8; one note on that item though, the standard currently only supports patterns of features of size at MMC, not at RFS. That, however, can be addressed / remedied in a corporate addendum ;~}

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[NormCrawford]

This is an exceptionally good thread. As can be expected, some really good input from Jim.
I do want to add a question for thought with regard to the co-datums as shown in the drawing. How is that any different then using a "pattern" of features of size as a datum feature? Where is the "origin" of a circular pattern of holes all of the same size and size tolerance (just to keep things simple). And with all due respect, does the origin of the nice circular pattern have to be shown on the drawing? No, it doesn't.
The point is that once the 6 degress of freedom have been restricted, the location of the origin can be anywhere. Of course, as I believe Jim mentioned, there becomes an issue when referencing a tertiary datum and just exactly what that may do.
I'm looking forward to this forum. Glad I found it.

Norm Crawford
GDTP-S
Applied Geometrics Inc.

http://www.gdandt.com

 

[DesignBiz]

The difference between a pattern of holes referenced a single datum and the JPEG of the tube shaded image, with 2 cylindrical features of size, at compound angles to each other (both ends of the tube run) each end having an axis and 2 planes 90 deg to each other, is:

For a hole pattern referenced as a datum, I could have lets say, 4 holes basically dimensioned 2.0" inches apart vertically and horizontally. Lets say these holes
are labeled datum B and a perpendicular surface that they intersect is labeled datum A. I can easily see that in a Feature Control Frame if primary is A and B is secondary, that I can construct a plane aligned with A and 2 planes at right angles to each other and that are perpendicular to A. The origin would be at the center of the hole pattern. All easy to create geometrically and fits the standard's definition for DRF (3 planes perpendicular to each other). Each datum reference block in the Feature Control Frame refers to a single datum. In the case of A, a plane. In the case of B, 2 planes 90 degrees to each other centered between the basic horizontal and vertical dimensions.

In the case of A-B as called out in the "drawing JPEG" (one thread prior to the shaded "image JPEG"), there are 2 features of size which in this case
are 2 cylinders that are at "compound angles" to each other. If a person tries to create a "single datum" from these "2 features", because they show up in one
datum reference block, it is not geometrically possible to create a single plane, axis, or exact point from the two of them, as the standard's definition of a datum would require. Nowhere that I can find in the standard does it state that a datum reference block in Feature Control Frame can be anything else than a single datum.


DesignBiz

"Quality is in the details"

 

[MechNorth]

Tks Norm. Good to hear from you.

Here's the link for a sample of what I've been discussing:

http://www.profileservices.ca/files/tidbits/thd1103_235809.pdf

As for the differences between multiple datum features & patterns of features of size, here's a big one; patterns of features of size must (currently, per Y14.5M-1994) be referenced at MMC, and thus a boundary method of control for related features is facilitated. In that case, one doesn't really care where the origin of measurement is because it's essentially / primarily expected to be a go/no-go gauge. For multiple datum features, there is no such MMC requirement, and therefore the datum features could be established RFS if beneficial to the design; hard-gaging is not possible in this case.

I've encountered situations where a pattern of features (RFS) is needed to establish the datum, so it was written into a corporate addendum or other supporting documentation. The origins of measurement were then documented on the drawing, similar to what I've shown for the tubing example.

DesignBiz, the only time that you get a single datum is when you have a single planar datum feature. As soon as you reference a cylindrical or conical feature, you get a datum axis which is the intersection of two mutually perpendicular datum planes. In the case of a spherical datum feature, you get a datum point, which is the intersection of three datum axes, each of which is the intersection of two mutually perpendicular datum planes. Thus, you conjecture that one datum reference means one datum is invalid. The number of datum planes, lines or points established by any individual datum reference within the DRF, is established by the nature of the datum feature, and the precedence in the datum reference frame.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[DesignBiz]

Jim,

I will take a look more closely in the morning. What you show seems understandable to me. If you look at the drawing that I posted above you will see a quite different drawing. One I dont understand nor can anyone here explain it to me. Nor can anyone tell me how it is inspected from our suppliers.

Would you possibly add a FCF callout to your example?

Thank you !

DesignBiz

"Quality is in the details"

 

[MechNorth]

DesignBiz, I quite appreciate your dilemma with your drawing. It needs to have the clarifying details added as I've done on mine. Also, the FCF of the general profile tolerance is inconsistent with the A-B datum used elsewhere. While it may be legal, it doesn't show a consistency of design intent.

There's a general surface profile in the title block of my drawing, which references A-B and no other datums as they're not needed. Adding secondary and tertiary datums would be possible, but then I'd have to exclude constrained degrees of freedom from the primary A-B. Another topic for another day. The general surface profile tolerance has extra boxes in the DRF because I haven't mastered SolidWorks yet, and can't get the DRF to update the size & number of cells automatically.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[NormCrawford]

If a pattern of FOS is irregular (like on an oil pan), where is the center of the pattern?

As for any pattern of features of size MUST be referenced at MMC, you got me. I honestly don't recall that requirement.

If Datum feature A, being a diameter, is correctly positioned with a FCF that references A-B where as B (a FOS)also references A-B all at RFS and IDEALLY are coaxial and as if the mating part is variable in how it grasps the two diameters to establish the Datum A-B rfs axis, isn't that a pattern? And isn't that a legal specification? I am not trying to be difficult! I know I am coming into this thread late and maybe it needs to be a new thread.
But alot of what I read I like, but some makes me wonder.

Norm

 

[MechNorth]

Norm, agree with the point you're making re irregular pattern of FOS establishing a datum; was a long term issue for me as well until I was shown to do what I usually refer to as "surrogate" datums / measurement origins, as I've done in the drawing. Prior to that, I'd been told a variety of silly things like find the center of area, center of mass, etc., and that there could be no origins of measurement and hard-gaging was mandatory.

Re pattern of features being at MMC, to more specific, the standard only shows and represents how to do datums based on patterns of features at MMC; patterns of features of size not being used to establish a datum presumably can be done at RFS. It doesn't specifically preclude RFS from use on datums established by patterns of FOS, but it is in no way communicated how to handle it RFS, and as I recall even the cert exam had a question on the MMC requirement.

I think you're right though, that the discussion of your last paragraph should be kicked into a new thread. I'll leave that up to you; pls add a graphic too. My brain is starting to hurt, and I like pretty pretty pictures!

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[NormCrawford]

Jim,
I understand the "surrogate" datum technique. Nice drawing. I see you do like pretty pictures and so do I. But you know how I feel about "drawings". Right?
Frankly though, I like that sort of drawing to be in the measurement plane and not on the part drawing. Just IMHO.
I agree with your clarification on MMC for patterns. I just wanted the group to be aware too. I'm always careful with the "must" word.
When I have time, I may start a new thread but basically, the last paragraph is simply discussing coaxial diameters that are far appart like on an axle or drive shaft.

Norm Crawford
GDTP-S
Applied Geometrics, Inc.

http://www.GDandT.com

 

[DesignBiz]

Jim,
I completely comprehend and agree with your descriptions of possible datums which can be derived from various features. I don’t have a problem with the DRF concept regarding what you have mentioned. Interestingly too, as you point out for a pattern referenced as a datum, it would require a material condition. For me under any of these scenarios you mention, in light of section 4 of the standard, I can apply basic geometry to agree with the callouts.
My continuing problem is to see the relationship of “multiple datum features” defined by 2 cylindrical features of size “at compound angles” to each other. They cannot meet the standards definition for a single datum geometrically. The big question to me at this time is how would you explain the DRF concept under these conditions? I understand the concept for a part to be constrained by both ends of the tube in this case and how the multiple datum concept captures all degrees of freedom, however it seems as if none of the standard’s DRF requirements apply to this multiple datum feature concept. What does one do with section 4 of the standard considering the multiple datum feature concept we have been discussing?
If you would read 4.1 in view of our multiple datum feature discussion, it appears quite obvious to me that this concept requires total disregard of section 4 of the standard (Datum referencing). Even section 4.2.2 regarding positioning the part refers to three mutually perpendicular planes to “position the part”.
I don’t think anyone can explain this better than you have. This concept of multiple datum feature part positioning seems to merit an entirely different section. Maybe some of you are smart enough to glean the” extension” of the DRF concept, however I am not. Simply comparing the drawing that I posted to any part of section 4 does not make sense to me. Even if a simple note such as the; “perfect form not required at MMC” and this case maybe “Multiple Datum Features position part; DRF does not apply”, would help me tremendously.
Thanks for your patience.


DesignBiz

"Quality is in the details"

 

[MechNorth]

No problem, DesignBiz. You're right that the standard doesn't explicitly cover this situation; I'm not sure if the next revision does either, but we'll see this year. After sittin in on the Y14.5-2009 meetings, it's clear that in some cases many extensions have been considered and are going to make it into the standard because they've reached consensus; in other cases, no consensus and not going into the book. Many, maybe even most of the extensions need extremely deep knowledge of the entire standard; that usually comes from senior level certification (GDTP-S), and even then it's not absolutely clear in some cases. You'll tend to see Dingy2, NormCrawford, and myself debate some items and corroborate on others; we're senior-level certified on the '94 standard, and we still remind each other of subtle things that easily slip our minds. Overall, if you're not comfortable with a solution proposed in this forum, don't use it; it's you that has to explain & defend it to your company. The easiest alternative in the tubing case is to select one cylindrical end as the primary datum feature which garners an axis with two planes, then select an end face as the secondary datum feature, gaining you the third datum plane. A surface on a bent segment can act as your tertiary datum which will arrest rotation.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[axym]

DesignBiz,

Again I applaud your efforts to make sense of this. Trust me, the multiple datum feature concept does agree with the DRF concepts of Section 4. I'm enjoying the challenge of trying to explain it. I'm glad to see that Norm has joined the forum, as I know that Norm "gets it" as does MechNorth. Hopefully between the three of us (and others)we can put it in a form that is understandable. Any understanding that I have of these issues has taken literally years of pondering, "aha" moments, picking apart and questioning the statements in Y14.5, and endless discussion and debate with "buddies".

I'll reiterate my opinion that part of the problem you're having lies in the 1.3.3 definition of "datum". Trying to make sense of datum referencing while taking 1.3.3 literally is difficult or impossible - believe me, I've tried. To me, the theoretically exact planes, lines, and points are only understandable for simple combinations of simple datum features. Even then, the whole idea of extracting datums from the TGC's has become a conceptual side trip for me - I don't see it as a useful bridge to get from TGC's to a datum reference frame.

The definition in 1.3.3 also states that "a datum is the origin from which the location or geometric characteristics of features of a part are established". This isn't generally true. It's only true in simple cases in which the datum happens to coincide with a DRF component such as a coordinate plane or axis. To me, the DRF itself provides the coordinate system in which measurements are taken.

More tomorrow,

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[DesignBiz]

Axym,

Thanks to your perserverence, as well as Jim and Norm, I maybe having an "aha" moment. Maybe a false alarm, let's see.

All along I have had trouble with the three mutually perpendicular plane construction from the 2 features of size at compound angles. I believe Jim made and interesting statement that the designer arbitrarily orients the planes as in the case of a cylinder.

If I can take the planes established by one cylinder and then can arbitrarily pick a "point" from the axis of the other referenced datum (eg A-B), now I can geometrically create the 3 mutually perpendicular planes of a DRF. Prior to this I was trying to do this with the combination of planes and axis of the combined datums. I had not been thinking of a point as a possibility.

I am getting on track here?

DesignBiz

"Quality is in the details"

 

[NormCrawford]

DesignBiz,

In short, yes! You are on track.

Norm Crawford
GDTP-S
Applied Geometrics, Inc.

http://www.GDandT.com

 

[MechNorth]

Yup, works for me.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[axym]

DesignBiz,

I thought about your idea of arbitrarily picking a point from the axis of the other datum, and it doesn't work for me. But if it works for Norm and Jim as well, perhaps I need to think about it further. I sometimes see things differently than others here because I don't come from a design background. I spent several years doing CMM programming, in which a primary task is to take imperfect features on a real part and construct datum reference frames around them. This means that I sometimes see things in a way that most designers don't, and sometimes miss things that are obvious to any designer!

The main problem I have with the arbitrary point datum is that the origin and clocking of the datum reference frame would also be arbitrary. For argument's sake, let's say that the first datum axis defines the Z axis of a coordinate system. This axis gives us the orientation of the XY plane but not its location (i.e. the Z origin). The first axis also gives us the intersection of the XZ and YZ planes but not their clocking. If the second axis is at compound angles to the first axis, then different points on it will result in different Z origins and different clocking. This just doesn't sit well. Am I missing something?

Establishing the first 2 datum planes from just one of the cylinders in the A-B reference doesn't sit well either.


Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[NormCrawford]

Evan,
The key is in simulatneous requirements. The "arbitrary" may be a bit confusing, but in reality, the origin can be as long as it remains the same for the other requirements referencing the same DRF. So, yes! You could have more then on DRF for inspection, but one or another has to be chosen and then stick with it during that particular inspection.
I have seen enough of your posts to know that you are familiar with all of this. I think some of the wording like "arbitrary" and "point datum" may be part of the problem here.

Norm Crawford
GDTP-S
Applied Geometrics, Inc.

http://www.GDandT.com