Another Angled Hole Positional Tolerance Question 4

[R1chJC]

Hi All,

I have a question regarding the use of positional tolerance on angled holes.

Is the attached legit?

Thanks,
Rich.

 

[greenimi]

I am not disagreeing....Well sort of.....

Just to stir the pot a bit: Fig 4-7 -2009 has a 150 basic dimension shown from something that is not a datum feature.
Also 7-55 has a 200 basic dimension to the point where the hole's "true position" (I don't like this term, but I am paraphrasing the book) axis intersects the surrounding face.

 

[axym]

Hi All,

I would maintain that the two Before and After dimensioning schemes are equally valid. I disagree with the textbook's description of the "error" in the Before figure - the presence (or absence) of basic dimensions does not affect what the datum features are or how the true position is established. I don't see it as an example of an implied datum.

The basic dimensions only apply to the theoretically exact geometry, and could be described as dimensions of the model. The basic dimensions need to define the theoretically exact relationship between the datum features and the toleranced features, which then defines the relationship between the datum feature simulators and the tolerance zones. The datum feature references determine how the actual part gets lined up to the datum feature simulators - and hence how the datum reference frame and tolerance zones are established on the actual part.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[CheckerHater]

The point of axis intersecting a surface is not a feature.

That's what makes "before" example meaningless and "after" acceptable. In fact, "after" will work even if you don't know where the surface in question is. That's what makes it robust.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[pylfrm]

Picking up on the points made, how about the attached scheme?

You need to define the basic angular relationship between datum features B and D, or between datum feature B and a hole axis.

Do you have some objection to dimensioning the distance between a hole axis and the intersection of datum features D and B? I realize there are limitless possibilities here, but that option certainly certainly has advantages.

If there is a mating part with a corresponding hole pattern, it could be helpful to make the dimensioning schemes match. Also, it might be a good idea to choose a scheme that doesn't involve non-terminating decimal dimensions which require rounding for display.


Regarding 2018-05-04_111730.jpg posted by SeasonLee:

The problem with the "Before" scheme is that the basic relationship between the hole axes and the datum features is not fully defined. The hole axes are dimensioned based on their intersections with the angled surface, but the basic relationship between that surface and the datum features is not fully defined, so the chain falls apart.

The "After" scheme solves this problem quite thoroughly by eliminating the position tolerance altogether.


pylfrm

 

[gabimo]

The "After" scheme solves this problem quite thoroughly by eliminating the position tolerance altogether.

That is NOT TRUE! Look in the original book and see that the position callout is still there. Not sure why SeasonLee removed it (position callout) intentionally or unintentionally.

SeasonLee,
Please, do not play tricks with us, as people are trying to justify your altered scheme as being valid or justify something that is not correct in the first place.

 

[SeasonLee]

@gabimot
The book I bought from Amazon is really dirty and old, so I can't snapshot the whole page and I did removed the dirty covered area intentionally, but forgot to add the position callout manually.

Sorry for any inconveniences caused.

Season

 

[gabimo]

Season,
No problem. Now you have at least one clean page in your book (my attachment above)

Interesting enough, how people are trying to rationalize the "altered" content and pass it as a recommended approach. Probably we, as a “common users” are trusting too much the books (written by SME’s, committee members, etc.), we are putting too much faith that what someone wrote is entirely correct and unambiguous. Maybe a “grain of salt” does not hurt here and there.

 

[axym]

greenimi,

Way to stir the muck ;^). I agree with your observations on the Y14.5 figures. They show that basic dimensions can be applied to any feature (even features that are not the considered feature or one of the datum features for that FCF) or even to things that are not features (imaginary intersection points out in space as in Fig. 4-7). This does not cause any issue because the basic dimensions are not applied directly to the actual part geometry.

CH,

I agree that the intersection point of an axis and a surface is not a feature. But it doesn't matter in this case - the basic dimensions don't need to apply to features (see above). So I would maintain that the Before dimensioning scheme, and the After scheme is not necessarily better. Another advantage of basic dimensions is that if you don't like the ones that were annotated on the drawing, then you can calculate others and use those instead. As long as the relationship between the considered feature(s) and the datum feature(s) is fully defined (which I don't think it is in either the Before or After scheme), then you can use a different scheme and it doesn't change anything (except possibly rounding-related inaccuracy as pylfrm mentioned).

gabimot,

A while ago (coincidentally, after I started attending the committee meetings) I learned not to blindly trust books written by committee members - they don't all agree, so they can't all be correct. I don't even trust everything in the standard itself ;^).

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[3DDave]

The problem with the 'common error' in the example is that as the length of the vertical portion changes the horizontal component of the angled surface changes, so the intersection of the axis with that surface moves horizontally leaving the location of the axes unfixed relative to [A|B|C].

In the 'one correct method' version, the location of the axes is fixed within the [A|B|C] without regard to intermediate geometry.

It seems like a driving principle of true position is that its location is fixed to a DRF.

 

[CheckerHater]

Evan,
There is all the difference in the world there.
"Before" example controls neither location or orientation of the holes.
So, you need 2 more coordinates to locate the points, 2 to 4 more angles to orient the axes, or make inclined surface datum, specify holes perpendicularity to the inclined surface, all to get to the clumsy, ambiguous equivalent of "After" set-up.
Ambiguous because it is not clear if intersection points are randomly hanging in space or belong to inclined surface. If they belong, then you don't need basic angle, as surface is already drawn thru 2 points, etc., etc.
And you still insist that "before" is actually equivalent or even better?

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[chez311]

I apologize for the lengthy response - I've been following this thread pretty closely and theres a lot to process here.

I really think that what this really boils down to is what is most clear and concise way of communicating all the important dimensions and design intent while providing a proper dimensioning scheme and datum reference frame so it can be inspected. I think we can all agree that there is more than one way to skin a cat, and some of us may prefer one way or another but I think its hard to argue that one way is truly right or wrong - as Evan has pointed out as long as all of the proper basic dimensions are there to fully define each feature then technically it may be acceptable no matter how convoluted.

Ambiguous because it is not clear if intersection points are randomly hanging in space or belong to inclined surface.

Convoluted or requiring additional dimensions to define something does not equal ambiguous. When I look at that drawing I see dimensions to points clearly at the intersection of the surface not hanging out in space - see my attachment, a dimension to a point like this is required for the "after" method if the datum is also an axis. I agree that the drawing shown by Drake is incompletely dimensioned (as is my example) however even with the "after" method presented as correct that surface is still floating even if the axis of the hole is fully defined - I think the assumption would have to be that any additional basic dimensions to define that axis would be added.

THAT BEING SAID - I think my gut reaction is to side with the "after" method presented by Drake as has been posted here multiple times (and I actually initially posted much earlier in the thread) as I agree that it is more robust. This is because it relies on fewer dimensions to fully define the location of that axis in space. Any time there is an opportunity to present dimensions more clearly and concisely is typically good - take the example of a flange with 12 bolt flange. Those 12 bolts could all be basically dimensioned with 24 x/y positions, the same result that could be accomplished with 2 dimensions (1 angle and 1 bolt circle diameter). Is each technically correct? Yes but one results in a huge reduction of the number of dimensions as well as the difficulty of interpreting the drawing. Now there could be other situations where the x/y positions would be preferred - say if some or all of the holes are not evenly distributed in the radial or angular directions around a bolt circle . Ultimately it would be up to the designer to decide which scheme, or combination of schemes, is best.

However not to add another variable into the mix but if you take a look at Figure 7-36 in the Y14.5 standard it actually presents another method to provide the dimension where the hole axis intersects the datum axis. See attached for my comparison between the methods presented by Drake and this method.

 

[CheckerHater]

chez311,
Your comparison is not valid.
Your "after" sketch is not equivalent of Drake's "after" sketch.
I am not planning to start lengthy discussion, but Y14.5 has certain rules on how to dimension feature of size and the way dimension is attached to the round hole (FOS) is incorrect.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[axym]

Hi All,

It seems that each of us is focusing on different details and their importance, but I think that we're generally in agreement.

CH,

OK I think I see it now. The Before scheme does not provide enough dimensions to fully define the required geometry, and the After scheme does. With only a Y coordinate for one endpoint of each axis, we don't know where the axes are in the X direction. If we wanted to build a gage to check the position tolerance, we couldn't do it using the dimensions in the Before drawing But we could build a gage using the After drawing. So I agree that from a completeness standpoint, the Before scheme is insufficient and the After scheme is sufficient. Sorry for not seeing that earlier - 3DDave pointed it out as well.

As far as what type of basic dimensions can be used, I would still maintain that the essential thing is that the relationship between the toleranced feature(s) and the datum feature(s) is fully defined. Many different arrangements of basic dimensions could accomplish this. But I would also agree that certain arrangements would be more direct and clear to the typical reader, and other arrangments would be less direct, unclear, and convoluted.

Here is an example. I've added extra dimensions to the Before scheme to make it complete (please forgive the poor drafting technique - I'm not an expert on drawing layout and drafting).

I would say that this dimensioning scheme is adequate, but perhaps somewhat indirect. The dimensions involve the left-hand end surface that is not a considered feature or a datum feature for the position tolerance, and this may be distracting or confusing to many readers. But including dimensions to this surface does not cause any fundamental issues or make the scheme meaningless, because the dimensions are all basic anyway. It doesn't mean that we have to line up on the left-hand end surface on the real part.


Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[chez311]

CheckerHater - not with the intent of dragging you into a lengthy discussion but for my own edification, could you at least point out what exactly is wrong with it so I can look into it further myself? You are referring to the basic dimension, not the size dimension correct? I'm not sure why the "before" sketch would be okay but the "after" would not as it is just its inverse. Or are they both invalid?

I am still learning and I would like to not make the same mistakes in the future.

 

[CheckerHater]

I am saying exactly the opposite - "before" does not look OK to me

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[chez311]

Your "after" sketch is not equivalent of Drake's "after" sketch.
I am not planning to start lengthy discussion, but Y14.5 has certain rules on how to dimension feature of size and the way dimension is attached to the round hole (FOS) is incorrect.

I understand you did not agree with the "before" sketch, in your response to me it seemed like you drew issue with how I applied dimensions in my "after" sketch and I was trying to clarify why.

 

[SeasonLee]

With only a Y coordinate for one endpoint of each axis, we don't know where the axes are in the X direction.

For my understanding, the basic dimension on the X-axis is implied zero, so I will think the features already fully defined, its not necessary to add extra dimensions.

I am always learning here, please let me know if I am wrong.

Season

 

[chez311]

SeasonLee - i think he was just talking about the x and y in the side view that he added the dimensions to, his 1.xxx, 1.yyy, and 1.zzz are all in the x direction he was referring to.

 

[3DDave]

Season Lee,

If the basic dimensions locate the X and Y of the intersections with the nominal 30 degree surface, don't they also set the slope of that same surface in a redundant manner? It would be better to pick some random locations on the axes that are far from the surface to establish their location. If the dimension was to the projected intersection with [A] then only the Y value would be required.

(Spelling or typing. Sigh.)

 

[R1chJC]

If the intent is imply the intersection point of the angled face and the holes, wouldn't it make sense to use that as a primary datum? Otherwise I would think that the full 'x/y' co-ordinates and angle are needed to determine the true hole axis as per SeasonLee's picture above.

I often wonder what difference it makes discussing the nuances of things like this. In my humble experience, if I ask 3 engineers I get 3 different answers. If I ask the Manufacturing Engineers or Inspectors I get another different set of interpretations. I seems that even the Standard is not clear at times.