Pattern of holes as a datum which is distributed only for a quarter of circle.

[Madhu454]

Please see the attached drawing.

Function of the part, 4 flange segments assembled to form a complete ring. Holes in the flange are clearance holes for bushings.

1) is it legal to use pattern of holes as a datum feature which is distributed only for a quarter of circle? Please see the attached figure; the holes are distributed only for 90 deg range.

2) Primary datum is A, secondary datum B is the axis of pattern of holes, which is the intersection of two planes at 90 degrees. The dovetail profile is located and oriented wrt to datum-A and datum B at M modifier.

There is a problem; the hole-hole location is controlled using basic dimension. But I need to locate the first hole in the flange to the end face of the dove tail profile. Is that Correct? Because the end face is not a datum feature. Also the dovetail profile is in line with the datum-B; hence I no need to locate it WRT datum B, which implies 0 basic dimension,

In this case how do I locate the first hole of the pattern?

 

[PaulJackson]

If the hole pattern via the bushings in the four quarter segments constrain the three degrees of freedom not constrained by the primary then yes it is not only legal but functional as well to declare the pattern as the secondary datum feature.

For the ends and radii just detail their basics relative to the pattern with a profile all around (if areas of the perimeter's profile need to be improved for tolerance do so with zones).

BTW - The secondary datum feature is not the axis the pattern but rather the pattern of axes.

Paul

 

[PaulJackson]

There is a difference between a pattern that functions to locate and orient related features in assembly and a pattern that simply must assemble where departures from MMB size assist in that assembly.

For the later a MMC tolerance modifier is functional but for the former permitting additional hole pattern location tolerance based on feature sizes may not be. A good check on the appropriate use of MMC is to ask "Does the feature's function degradate as it is permitted to deviate from its ideal location or orientation" If the answer is yes then applying a MMC tolerance may not be functional except in some moderate way to acknowlege fit.

Paul

 

[Madhu454]

Hi Paul,

I did not get the meaning of the below in your post.

"BTW - The secondary datum feature is not the axis the pattern but rather the pattern of axes."

Does it mean axis of the any hole in the pattern can be taken as datum? or it is the PCD formed by all the holes? really I did not get.

Please help in understanding this.

 

[fsincox]

PCD? not sure what that is, it is all the holes as currently stated on the drawing.

 

[Belanger]

The secondary datum feature is not the axis the pattern but rather the pattern of axes.

Not to nit-pick... It is actually OK to say that the secondary datum is the axis of the pattern. (The datum feature would be the four actual holes themselves, not axes.)

See Fig. 4-26 of the current standard. Also compare the wording in paragraph 4.12.3 of 2009 to that of para. 4.5.8 in 1994. I suspect this was a conscious change, in order to place more emphasis on the datum feature simulator.


John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[fsincox]

kind of curious how others feel about the non-basic dimensions defining the are from pattern center? and is it primary or is it secondary?

 

[fsincox]

sorry, arc!

 

[pmarc]

1. I am with J-P in saying that datum B is a single axis.

2. Frank, yes, those arc should most likely be basic and controlled by profile of surface callouts. For arcs less than 90 degrees it is really hard to find arc's center (and in consequence size of arc) in repeatable manner.

3. Madhu, which standard governs the print? Is it ISO as in majority of your questions or ASME this time? I am asking because if ISO, I do not think you can go with (U) modifier followed by 0 tolerance value. First of all ISO 1101:2012 defines its own modifier for that - UZ, and secondly 0 would make no sense because this value in ISO defines offset between theoretically exact profile and CENTER of profile tolerance zone. Zero would mean equal distribution of profile tolerance zone relative to theoretically exact profile, so would be equal to a profile callout without any modifier.

 

[axym]

Hi All,

I would say that calling Datum B a single axis is okay per the standard but it's wrong ;^). I agree with J-P that the Y14.5-2009 shows the single axis in Figure 4-26, but I think the figure is misleading and confusing. That figure is one of my biggest pet peeves in the entire book. I agree with Paul that the datum is really not the "axis of the pattern", because such an axis cannot generally be defined on real part geometry. The as-produced part in the "means this" figure has holes that are at LMC size and perfectly located, which is a very special case in which an "axis of the pattern" could be uniquely defined. But this is not possible on typical as-produced patterns that have some relative location error and imperfect pattern shape.

Describing the datum as a single axis also leads to the incorrect conclusion that the DRF can rotate about that axis. The pattern constrains all three remaining degrees of freedom, including the last rotational degree. If the datum is described as a single axis, it is very difficult (for me, anyway) to explain the distinction between a secondary datum hole pattern and a secondary datum single hole. If we must define a datum from the pattern, it would be a "line on a plane".

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pmarc]

Evan,
As for first part of your post, since datum axis B is derived from 4 datum feature simulators of a certain size which are perfectly located to each other and not from the datum features itself, I am having troubles to see your concern. I mean, I do not really understand why do you think that actual configuration of holes in as-produced part and their size has any impact on how datum axis (or however we call it) is established. Am I missing something? I think fig. 4-26 shows the holes at LMC simply to emphasize the effect of datum shift described in 4.11.9.

 

[Belanger]

Right, pmarc. Evan, you wrote that:

...the datum is really not the "axis of the pattern", because such an axis cannot generally be defined on real part geometry.

But we never said that the axis is defined on the real part geometry. The axis is defined from the datum feature simulator. Notice the difference in those statements.

I didn't mean to sidetrack the discussion, but when Madhu asked the clarifying question, I made myself look at the actual wording of the standard.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[fsincox]

Is the axis of the pattern and the centroid of the pattern the same or are you refering to the basic arc axis, here? I sometimes have used the tern "the centroid of the pattern" but that is not really the same here.

 

[axym]

pmarc and John-Paul,

Rereading my earlier post, I've mixed two different issues together into one confusing mess. You're right that the datum is defined in the datum feature simulators and not in the actual holes. The standard doesn't say otherwise, and you two didn't either. But I still have issues, so I will try to articulate them more clearly:

First, I still disagree with the description of the secondary datum in Figure 4-26 as a "datum axis". To me, this description doesn't capture the fact that the hole pattern constrains the third rotational DOF. If the secondary datum feature was a single hole, the datum would be a single axis and the third rotational DOF would still be open. So I would say that to describe the secondary datum from a pattern of holes as a single axis is misleading at best. How is this axis different, and how does it constrain the third rotational DOF? Interestingly, the datum axis is only mentioned in the figure and is not mentioned in the text in 4.12.3.

Second, I would still say that there are aspects of Figure 4-26 that are very misleading. The planes of the datum reference frame (and an axis at their intersection) are shown, which is fine because these are derived from the perfect simulators as you clarified. But other planes (and an axis) are shown as well, which were derived from the actual hole pattern. This is the part that I really have an issue with, as well as the indication of the "displacement of the hole pattern with respect to the datum reference frame". The figure heavily implies that we can extract a unique axis from the actual hole pattern, and use this axis to evaluate the displacement of the hole pattern with respect to the DRF. This axis is not uniquely defined on real as-produced patterns in which the holes are not basically located relative to each other. I would also say that we shouldn't even try to evaluate the displacement of the "hole pattern" in the first place - we can only evaluate the displacement of each hole.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[fsincox]

The problem is how do I explain to people a way to check a part like this if I don't have a gage. I am happy we all have accepted this approach, but how do we ask someone to check it? Can we paper gage it?

 

[fsincox]

Al Neuman used to teach a method for "paper gaging" hole patterns that is why I refer to the centroid because if you lay all of the holes on top of each other you can get a centroid of the pattern.

 

[Belanger]

Thanks for clarifying, Evan, and yes I did notice your earlier point about an axis itself not stopping rotation. I just didn't reply on that point cuz I agree It's not blatantly wrong to say it's an axis, but it'd be better to say that it's an axis with clocking control or some such statement.

As to Frank's point about a centroid... Assuming we're talking about the same idea (a mathematically derived "center of gravity" of the pattern), that seems troublesome to me, although I have heard some explain it that way. Think of a half-dozen holes scatterned into a randomly shaped pattern. Sure, a CMM could find a centroid of any weird shape, but what value does that add? There's not functional use for that centroid, nor would any dimensions originate from there. I think the paper gaging idea is fine; maybe we're using the word centroid in different ways.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

http://www.gdtseminars.com

 

[Madhu454]

Hi All,
Thanks for your feedback.
By going through the above posts I understood this way, Please correct me if I am wrong.

The Datum axis B is not a single axis (axis of the pattern), it is the axes of VCB located at its true position. All these axes will altogether form a Datum Reference Frame which constrains the remaining 3 DOF. But little confused about how to derive the datum reference frame and where to locate it.

The example shown in the standard is little easy, I mean the datum reference frame is created at the center of the pattern (axes of 4 VCB located at true position). Some patterns may be such that you may not be able to derive the axis of the pattern. See the attached drawing, in which the hole pattern consists of 12X holes as shown. The question is what is the center of the pattern? Where do I originate the DRF? Also I have another hole which needs to be located to datum feature-B. To which hole do I need to locate that? As per my understanding I can locate the hole to datum feature-B for any of the hole in the pattern. The origin of DRF originates from that particular hole but (DOF is locked for rotation by pattern).

If I go back to my original posted questions. I found the answer this way
1) Yes, it is legal to use the pattern of hole as a datum, which is distributed only for the quarter of the circle.
2) I can call a basic angular dimension from the first hole in the pattern to the dove tail surface, and dove tail profile can be defined with Profile of a surface. I have tried to draw the datum reference frame for the flange segment. Please see the attached drawing.

Please let me know if my understanding is correct.

 

[fsincox]

J-P,
It is more the concept to remember that it is not one single hole or axis, exactly what you guys are discussing. I am just trying, like I suspect Madhu is also, to find a way to explain and calculate the center of the pattern of holes, its centroid if you will. CMM's can do it, mathematicians can do it, but can we make it practical so a guy on the floor can do it (with out making gages).