Multiple Bore Features as a Single Datum 1

[resjsu]

Hello,
So as topic states, 1 of our customer drawings has 4 holes called out as a single datum. Nothing I have read in the ASMEY14.5 standard (I could be missing it) really touches on this issue.

I have basic questions; is it allowed? How would you take dimensions using a datum composed of multiple axis etc.

If anyone has a link or some supporting information on this issue itd be appreciated.

I can post a pic if it helps but its pretty straight forward.

Thank you

 

[pmarc]

resjsu,

Do you have access to Y14.5? If yes, is fig. 4-26 in '09 edition similar to your application (4-22 is corresponding figure in '94 version)?

 

[resjsu]

Pmarc,
Excellent, I found it in the spec. So it is obviously allowed now in terms of measuring with faro/cmm since there is no pattern callout is it just treated as if the holes were individual? In other words if I have a 3 hole pattern that the customer is calling collectively datum -B-, when I measure on faro/cmm I measure each hole individually and just apply the position callout to each hole?

Maybe what I am struggling with is that if we use the datums to establish our perfect surface then how do I use multiple axis when trying to take measurements. It seems as though id need to average out the 3 datum axis to get an average point from which to take dimensions. Does my question make sense?

 

[pmarc]

resjsu,

I am not sure I have a full picture of your application. Are the holes in the datum pattern controlled with a position tolerance (to a primary datum) just like it is shown in the standard? At what material boundary is the datum pattern referenced in other geometric callouts - is it at MMB (M modifier after letter B in the feature control frame) or RMB (no modifier after B)? In one place you said that 4 holes are "called out as a single datum". In another you are asking what if you have a 3 hole pattern called collectively B. So how many holes are there?

 

[resjsu]

Pmarc,
I apologize. Both of the 4 and 3 holes was just an example to get a better understanding.. In either case my dilemma is the same, I cannot think of how to check multiple holes being used as a single datum on a cmm/faro. See attached PDF, it is a redline from our customer who wants to make the bolt hole pattern a datum.

 

[mkcski]

resjsu: Are the (10) holes in the same plane - axes parallel to each other?

Certified Sr. GD&T Professional

 

[resjsu]

Mkcski,
Yes, all 10 axis are parallel to each other.

 

[mkcski]

resjsu:

OK. So...the (10) holes will control (restricts) (5) of the (6) degrees of freedom (DOF) of motion. If datums B and C are on the side of the part, they are redundant/unnecessary/meaningless because datum A already controls those motions. The only motion remaining is "in and out" of the holes (perpendicular to their axes)

Certified Sr. GD&T Professional

 

[mkcski]

resjsu:

Sorry. Brain fart here. Ignore my last post. The (5) motions would only apply where datum D is the primary datum (in some other FCF).

Certified Sr. GD&T Professional

 

[pmarc]

resjsu,

There are two things to take care here:

1. Verification if all ten holes are within position tolerance relative to A|B|C.
For that you do not bother whether the group of holes have been defined as datum pattern or not. All you have to do is just check position of each hole in a normal way.

2. Establishing datum D from the pattern of 10 holes.
For that I am not able to answer your question because still some details about the presented dimensioning and tolerancing scheme are missing. So I will ask again: At what material boundary did the customer want the datum pattern D to be referenced in other geometric callouts (that we unfortunately do not see at the moment) - at MMB or RMB? Without knowing this I am really not able to tell you how to simulate datum D.

 

[3DDave]

I like the idea of using RMB on patterns like this. Pretty much every installation of multiple fasteners in hole patterns like this have expanding bolts that are simultaneously expanded until they just make contact in order to simulate the way software analysis of CMM results do. </sarcasm>

Realistically, if the CMM software won't handle this case, then send the measurements back to a CAD system user to model the parts to compare with a CAD model of a gage. It may take some fiddling to ensure there isn't interference between the cAD Gage and the CAD model, particularly in the MMB/MMC case. The underlying difficulty is that there can be infinite numbers of solutions to MMB/MMC alignments.

 

[resjsu]

Pmarc,
This is a redlined print that the customer is considering. On our side, we are trying to determine if we would be able to check the part if they made these changes. For educational sake could you provide basic info on how RMB or MMB would affect the datum scheme?

 

[pmarc]

resjsu,

The easiest and most common way of explaining the difference between calling out a datum pattern of holes at RMB or MMB is to use the concept of physical datum feature simulators (e.g. gage pins).

In your application, in both cases - RMB and MMB - the datum axis D would be derived from 10 gage pins and by default would be located at the geometrical center of the pattern of pins.

The difference between RMB and MMB is that in RMB case the pins would have to simultaneously expand until each of them achieves full contact with corresponding hole, while in MMB case the pins would have to be fixed in size (diameter). If in MMB case the datum pattern D would be assigned primary in some other geometric callout(s), that would mean the diameter of each gage pin would have to be equal to the MMC size of the holes, which is 17.4 (17.5-0.1).

One of the conclusions coming out of this description is that in RMB case the inspected part would not be able to shift (in terms of rotation and translation) relative to the pattern of gage pins because there would no physical loose between the holes and pins. In MMB case, however, the shift could be possible and it could be used as an advantage during inspection of other features controlled relative to D. This effect is commonly called the datum shift (take a closer look at "Means this" portion of fig. 4-26 to see what I mean).

Another conclusion/question coming out of the description should be this. Since the process of immobilization of the part for inspection should as closely as possible reflect part's interaction with mating components in a real assembly (this commonly called the function of the part), does it really make sense to derive datum axis from pattern of 10 features at RMB? This is what 3DDave pointed out with a bit of sarcasm. Is the inspected part installed in assembly using some expandable bolts that in addition expand simultaneously? Or maybe the inspected part is press fit with 10 pins/studs in the mating part? I don't think so. MMB seems to be much more rational choice in most cases.

 

[Tunalover]

Does this mean that inspector must use the basic hole pattern centroid and that all measurements must be made from there?

Tunalover
Electro-Mechanical Product Development
UMD 1984
UCF 1993

 

[pmarc]

tunalover,

No, this does not mean that inspector must use the basic hole pattern centroid as origin of measurements. I intentionally said "by default" ... :
a) ... because I did not want to make the post too complicated.
b) ... to imply that what I said could be overriden. See the statement in para. 4.12.3 starting from: "The origin of the datum reference [...]"

 

[Tunalover]

My apologies pmarc.


Tunalover
Electro-Mechanical Product Development
UMD 1984
UCF 1993

 

[pmarc]

No need to apologize, tunalover.

 

[resjsu]

Ok, so assuming MMB how do you find the "geometric center" of a 10 hole pattern? I see the example in the ASME14.5 but they are using 4 holes which is a little more intuitive. If I do not have to find the center, how then do I use this 10 hole pattern as a datum used to check other features? I need to be create well defined planes or measurements in my program so I can check other features to D, this is where I am having the most difficulty understanding the 10 hole pattern as a single datum. Gauging it makes a little more sense but checking with software is still very unclear.

 

[Belanger]

resjsu -- here's a similar thread that might help with your latest question:

http://www.eng-tips.com/viewthread.cfm?qid=381821

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

 

[resjsu]

Belanger,
Thank you, it helped reading some varying points of view. However, it didnt do a whole lot in the way of clarifying too much. I did key in to your response regarding taking measurements from basic dimensions relative to the datum. So, in my case after checking each hole individually against its TP, I will clock the part using 2 intersecting planes where the planes are established using pairs of holes. These planes will be used to check my part. This seems pretty simple but its still a large leap of faith to say with confidence that I am checking the part "correctly." So in summary, I will not find the center of the pattern but using a couple basic dimensions relative to datum D will create intersecting planes that will then simulate my datum D and use these planes to check the features called back to D.

If this is so far off the mark let me know but if it is "good enough" I will run with this.

Thanks again