Tolerance of Positoin - feature positioned to itself ??? 2

[DesignBiz]

Please review attached drawing for comment on the tolerance of position callouts circled.

1. I do not understand how a FCF can reference the feature positioned to itself in a datum reference block. (e.g. datum A is positioned with reference to itself in A-B)

2. I believe the location tolerance and the basic dimension relavent to it should have the same number of decimal places. (Not shown on the drawing) (e.g. if tolerance in FCF is .XX then its basic dimensions should be XXX.XX NOT XXX) Is this in the standard?

DesignBiz

"Quality is in the details"

 

[ewh]

<start sarcasm>
OK, so I guess they just shut the line down and not ship parts until everyone involved understands this issue.
<end sarcasm>

assuming that the parts are being inspected, and inspected correctly[//quote]
Isn't that the situation that exists currently with no one able to explain the callouts?

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[ringster]

If they fit at N/A and are interchngeable, you are home free.

If not, back to the drawing, excuse me, monitor screen.

Have you consicered Model Data Set definition for the parts..

 

[DesignBiz]

Ringster,

The “extensions” I refer to were presented by the senior GDT certified contributors on this site. (Ref other thread. As of now I do not know how to link that thread to this reply)

I agree that the authority for acceptance of an extension to the standard’s fundamental principles is based on the individuals credentials presenting the extension.
What may be an extension to one, may not make the muster for another.

Part inspection? Who knows if they are being inspected “correctly” !!! Never have seen anything like this before. Even the tubing in-house standard makes a comment regarding the amount of scrap and cut and weld required for tubes. I wonder if the lack of proper inspection could stop much of the scrap?

So as hard is it may be to imagine, no one here can say for sure how the parts are inspected. Somehow they are inspected and accepted and it is acknowledged that there will be considerable scrap.
I have worked on many tubes for gas turbine flight and power generation engines. We did not operate
like this at those companies.
Sorry, I am not making this up and it is a Fortune 500 company. Enough said?


DesignBiz

"Quality is in the details"

 

[axym]

It looks like we're getting into some philosophical issues here, like accommodating the limited GD&T knowledge of the people interpreting the drawings. It reminds me of something I saw on one of those "funny rules of life" lists:

"An easily understood, workable falsehood is more useful than a complex, incomprehensible truth".

I think that part of the problem lies in the reliance on the GD&T drawing itself. I often encourage designers to prepare other documents to go along with their GD&T drawings - sketches, "means this" diagrams, written explanations, etc. For the sole purpose of providing another means of communicating the intent for those who are not GD&T experts. This can be particularly useful for datum referencing. As we've seen from the "2 skewed cylinders" example, the GD&T can be complex for a relatively simple functional situation. A quick sketch of how a real part would be held in a fixture would probably help a lot in this case.

Another part of the problem (IMHO) is the common requirement to make GD&T drawings that people with average GD&T knowledge (or less) can understand. I see this all the time, where a designer has had to "dumb down" the GD&T because the downstream folks couldn't understand it. This is more a management issue than anything - someone bending tubes on the shop floor shouldn't be expected to be a GD&T expert as well. There is a real need for more "manufacturing engineering" people that can decode a GD&T drawing and convert it into manufacturing and inspection procedures that others can follow. But it seems that many companies don't have this luxury.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[ewh]

I totally agree, Evan, but that is the reality of the world many of us work in. I like the idea of generating additional documentation explaining any methodology used that might not be apparent just by following the drawing. In many locations (here for example) the documentation system may not be set up to accomodate this, in which case you could add it to the drawing in some manner (general note, additional sheet, etc).

"Good to know you got shoes to wear when you find the floor." - [small]Robert Hunter[/small]

 

[drawoh]

axym,

"An easily understood, workable falsehood is more useful than a complex, incomprehensible truth".

There was an article in the Skeptical Inquirer by Isaac Asimov entitled "The Relativity of Wrong". We make assumptions about stuff. Science comes up with a better, more accurate model. This usually does not change the fact that the original assumption was an adequate approximation for people trying to do practical stuff. For example, over short distances, the earth is approximately flat. You do not account for the curvature when you build a floor, or draw a city map.

Being confused and wrong is something different, entirely. You did day "workable", but I think "falsehood" is a little strong.

JHG

 

[PeterStock]

I had a question I posted on this thread earlier:

What would be the difference in interpretation of using
[A]B]C] as he suggests and using [A-B]C] as shown on the drawing?

I read the spec over and I may have answered my own question.

To determine the Datum Reference Frame on most parts [A]B]C] your would pick up 3 points on A, 2 points on B and 1 point on C. For the drawing referenced in the original posting, this would be inserting the tube in a datum reference for A, clocking the tube by referencing B and make contact on C.

For the [A-B]C] datum call-out I would insert the tube in
a datum reference for A and one for B at about the same time, then make contact on C. The first five points would be distributed among datums A and B.

As I see it a [A-B]C] datum call-out could be thought of as being met if either [A]B]C] or A]C] is satisfied.


Peter Stockhausen
Senior Design Analyst (Checker)
Infotech Aerospace Services

 

[drawoh]

PeterStock,

The weird thing about this discussion is I claimed that I have used the datum [A-B] a couple of times, and since then I have used it twice. In both cases, it was for a flat surface broken up by a slot.

I still don't see the point of datum[&nbsp;][A-B][C] on the OP's drawing. This is not an appropriate application for this. If you call up everything with respect to datums[&nbsp;][A], you would have located the part completely, and accounted for some of the flexibility. Even if the two datums are of equal importance, you need to fixture one of them first. All datum[&nbsp;]C can do is control more flexibility.

The sloppyness of the parts does not matter either. Call up loose tolerances.

JHG

 

[DesignBiz]

I really like this website! It is very very interesting.
Thank you drawoh ! My point from the beginning. In this instance the A-B callout is unnecessary from my point of view. As you mention in a surface broken by a slot or a rotating shaft that may use 2 coaxial bearing surfaces to define a single axis, all make perfect sense to me. Multiple features used to define a single datum. What troubles me with the multiple datum concept as an extension of the DRF definition is just that, multiple datums in one datum reference block. It is has been said that we can go back to the 1982 standard to carry over these principles (I don’t have a 1982 standard). It is confusing at best and in the case of symmetry if you went back to the 1973 standard it is a different meaning. Can we really rely on a previous standard to interpret the current one? I believe I understand the general concept that others have presented as to how the A-B constrains all the DOFs, however it does not meet the definitions in the 1994 standard. (e.g. each datum ref block in a FCF represents a single datum). I simply say documentation describing the new meaning in this case must accompany the callout declaring that the A-B is more than a single datum. I know some of you senior certified guys maybe frustrated with me, but that's my story and I'm stickin' to it !!!



Can anyone tell me what the difference of an A-B vs. A|B callout fixture setup would be? and what advantage there would be in this case (2 skewed cylindrical axes) of one over another?


DesignBiz

"Quality is in the details"

 

[ringster]

My opinion: as long as A aand B remain as shown, an AXIS of a cylinder at each end, neither A_B nor A/B will work.

 

[axym]

I've created a fictitious part that is loosely based on the one in the original post.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

 

[axym]

Disregard my previous post. I clicked the "Submit Post" button by accident. It looks like only one file can be uploaded in each post.

I've created a fictitious part that is loosely based on the one in the original post. Please see the file named nominal.jpg. Datum features A and B are the hatched areas at each end of the part. The nominal centerline is also shown.

The other 3 files depict an imperfect actual part, being fixtured in 3 different ways according to 3 different datum referencing schemes. Simulators A and B are basically located relative to each other, and the jaws are symmetric about the nominal centerlines at the datum areas. I deliberately left out datum feature C, as it is not necessary since A and B constrain all 6 degrees of freedom. I also deliberately left out anything to do with datums and datum reference frames - we can deal with that separately.

The file A-B.jpg shows the part in the controversial A-B scheme. The idea is that both simulators have been tightened simultaneously and at the same rate. Neither simulator aligns fully to its datum feature. The part has problems in the middle, but does not have really large deviations anywhere.

The file AB.jpg shows the part location if A was referenced as primary and B secondary. The A simulator has been tightened first, fully aligning with its datum feature. Simulator B was tightened second, and only touches its datum feature on one side. The position of the part is very different from that of the A-B scheme. The requirement to align to the relatively short datum feature magnifies the deviations at the other end.

The file BA.jpg shows the situation with B as primary and A as secondary. The position of the part is different again, with the deviations magnified the other way.

One key assumption in all of this is that the part remains rigid and is not deformed to conform to the fixture. This is the default condition in Y14.5, and applies unless the drawing has a note describing the restraint condition. Based on some of the recent posts, the real parts are most likely non-rigid and are being deformed to some degree when fixtured. This might explain some of the assertions that there is little or no difference between AB, BA and A-B - the parts are being deformed the same way in each case. However, if the parts were being inspected in the free state using a CMM, these datum referencing issues would be very significant.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

 

[axym]

Here's the AB.jpg file.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

 

[axym]

Here's the BA.jpg file.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

 

[axym]

Here's the A-B.jpg file.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca

 

[KENAT]

AXYM, you can upload more than one file at a time. However instead of using "click here to insert your file" or whatever it says, instead cut and paste the different links into you post.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

 

[DesignBiz]

Thank you Evan,

Your time to respond is appreciated. Most times for me,
graphics supporting the words can make things more clear.

DesignBiz

"Quality is in the details"

 

[ringster]

AXYM,

If Im reading the A-B jpg file correctly, it appears that it lacking in definition. To establish the simulators A and B, there must be a relationship between the two of them. As I view the sketch there is none. Would you not have to have a distance somehow defining the separation of the two plus an angle? Can you somehow clarify the sketch, to make the A-B DOABLE?

( my opinion; can't relate the 2 axes}

 

[MechNorth]

Good reply, Evan. Ringster, you would indeed need basic dimensions (linear distances & angles) to locate the two datum fetures wrt each other. That establishes the theoretically perfect relationship between the datum simulators as well.

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

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[axym]

KENAT,

I thought I had seen posts with multiple links! Thanks for the tip.

DesignBiz,

There are a lot of things in GD&T that I find hard to explain in words. This was obviously one of them - my earlier attempts to explain the difference between A-B and AB in words didn't get there. A diagram always seems to cut through. In fact, I usually learn something when making the diagram.

Ringster,

You're right, there is a basic relationship between the two simulators. They have a basic angle between them and a basic distance apart, and these were not shown on the diagram. I don't actually know what they are - the simulator centerlines were just extracted from the nominal CAD geometry of the part.

OK, I've annotated some dimensions on the nominal part that should allow you to reproduce the relationship of the simulators. Please forgive the drawing - I'm no detailer and the software I'm using isn't really designed for drafting. The dimensions and angles are of course supposed to be basic, despite the lack of boxes. The datum areas have a length of 3.

Evan Janeshewski

Axymetrix Quality Engineering Inc.
www.axymetrix.ca