Please help clarify datum shift vs bonus tolerance

[ozzkoz]

Hi,
I read a thread on here regarding datum shift and read through the section in the ASME standard. My understanding is this:

When the MMC modifier is applied to the tolerance in the FCF then bonus tolerance is allowed. If the feature is produced at LMC then the tolerance zone is increased by the difference between LMC and MMC. This is very clear.

When the MMB modifier is applied to a datum (secondary for example) then the MMB datum feature simulator is used to define the datum plane. If the datum was a hole, then you would take a MMB cylinder and place it in the bore and take measurements relative to the cylinders axis. If the hole was made to LMC then the part could shift relative to the datum feature simulator. I still think I am pretty clear to here.

When you check a hole pattern to the MMB datum, I do not think you get bonus tolerance for the difference between MMB and LMB datum feature. I DO think that you are free to adjust the part within the allowance between the datum and MMB datum simulator before inspection, and then once you decide on its position, you make ALL measurements required relative to this datum without re-adjusting the part.

If it is treated as bonus tolerance then I don't see how the fixed faster formula would work given that it doesn't account for bonus tolerance due to datum shift.

Is my understanding correct?

Thanks

 

[axym]

ozzkoz,

Your understanding sounds very thorough and correct to me.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[dingy2]

ozzkoz:

The bonus tolerance on the feature that you expressed seems correct. It is the difference between the actual size and the MMC of the feature. Right on.

As far as datum shift - mmmmmmmm. It now depends upon whether we are talking about ASME Y15.5M-94 or 2009 edition.

The 2009 edition uses the term MMB - maximum material boundary but not the 94 edition.

I am a bit confused about the use of MMB on a datum plane? One does not gain any bonus or shift on a datum plane even it it was qualified with a feature control frame using perpendicularity, as an example.

94 Standard - If one had a hole used as a datum and it was referenced in the feature control frame at MMC, the datum shift is the difference between the actual size and its MMC size (smallest allowable size).

If the datum feature was qualified in some manner such as perpendicularity at MMC, then the datum shift is the difference between the actual size and its virtual conditon size. On a gauge, one would use a virtual condition pin rather than its MMC size pin.

2009 edition - Now it depends upon whether the datum is primary, secondary or tertiary.

Let's look at the hole as referenced as a primary at MMC. It does not matter whether the hole was controlled or qualified with a feature control frame and has a virtual condtion size, the datum shift is the difference between the actual size and it MMC size.

If the hole was a secondary and it has been qualified with both a positional tolerance and a refinement tolerance of some sort. Then the datum shift is the difference between the actual size and the virtual condition size developed by the refinement tolerance.

If the same hole was a tertiary, the the shift is the difference between the actual size and the virtual condition size developed through the positional tolerance rather than the refinement.

2009 edition get a bit complicated.

Hope this helps.

Dave D.

http://www.qmsi.ca

 

[ozzkoz]

I'll have to check out 1994 but regardless of the edition, am I right in assuming that the datum shift does not become bonus tolerance?

If it was, then suppose I had a part with pilot hole, and two bolt holes (all three holes in a line). Let the bolt holes be controlled to the pilot at MMC, then if the pilot is at LMC and the measured shift is applied as a bonus tol on the holes then each hole could drift away from each other. A mating part might then not fit because the span between bolt holes is too much, yet I've seen the pattern toleranced this way many times.

 

[Belanger]

Dave -- I didn't know anything had changed regarding this in the 2009 standard, other than the terminology of MMB over MMC (which was misleading because it had always meant virtual condition anyway). What's the business about it depending on primary/2ndary/tertiary?

Also, I wonder about a hole modified with MMB being used as a primary datum. If there is a secondary datum, then the looseness caused by the primary's shift will actually make it contact the secondary datum feature simulator better. This kinda makes the secondary datum take precedence! (Read that again, but I'm saying that a primary datum can't have an MMB modifier unless it's the only datum. Thoughts, anyone?)

Ozzkoz -- datum shift and bonus do not add together to give you one grand tolerance number. This is because bonus tolerance allows the tolerance zone to grow. The shift tolerance doesn't allow growth, but it allows the zone to move around.

So in your example of the three holes (one being a pilot) the looseness around that pilot (which we call shift) is not added to each of the bolt holes' tolerances. Rather, they can both drift in the same direction a little extra due to the shift. They cannot drift away from each other due to the shift.

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[ozzkoz]

Thanks, this is what I thought, but how is this inspected?

I suppose it can be done a few ways, one of which would be to fix the part, find the location of both holes, plot them and then see if both holes have drifted by the amount allowed by datum shift. Alternatively I think if the holes are within tolerance you should be able to shift the part around to a point at which both holes are measured to be in the tolerance zone wrt the datum planes.

 

[KENAT]

Would I be correct in thinking you get an effect comparable to that from composite position? The holes are still controlled to one another the same, but the pattern can move a bit more?

Posting guidelines faq731-376

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

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[Belanger]

Hmm -- I guess it's sort of analogous to composite...

As for gaging, the easy answer is to use a functional gage. If the part fits on the gage, it's good. No math involved!

But yes, of the two choices mentioned by ozzkoz, the second one is probably best: "Alternatively I think if the holes are within tolerance you should be able to shift the part around to a point at which both holes are measured to be in the tolerance zone wrt the datum planes."

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[dingy2]

John-Paul:

There are massive changes relative to datums in the 2009 edition of ASME Y14.5-2009 beginning on page 48. I had to read it over many times and then I developed a web page for some of the main changes at

http://www.qmsi.ca/datums2009.html.

It has become very important to note the ASME Y14.5 level on the drawing.

Dave D.

http://www.qmsi.ca

 

[Belanger]

Thanks Dave! Now I know what you were referring to. I am aware of the changes; it just didn't ring a bell at first.

What threw me off is that the stuff about a datum referenced at MMC being primary, secondary, or tertiary didn't change in 2009.

I think it was always true (see 1994, para. 4.5.4 through 4.5.6). What they changed was just a more detailed explanation and they added the picture (Fig 4-16). Also see paragraphs A.6.8 and A.6.12 of the 2009 edition's appendix. But this might be good fodder for discussion.

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[Belanger]

All this brings me back to another thought that I had expressed above. In my previous post (6 posts above) I questioned whether a feature of size can have the MMB modifier if it is primary (followed by a secondary).

The current standard shows such a case in Fig 4-21(c). So I guess it's legal. My hesitation is this: as soon as the pin/boss is made to a slightly smaller size, the flange will be able flatten out on the top, as in Fig 4-21(b). Now it's back to being just like Fig. 4-20(c).

I'm just thinking out loud... But this doesn't really seem practical, does it?

John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[axym]

ozzkoz,

You've got the idea exactly right, so don't let our ramblings confuse you!

Dave,

I didn't think anything had changed in '09 with regards to datum shift, other than the use of the terms RMB, MMB, and LMB. It still works the same way that it did in '94.

I've always been uncomfortable with datum shift being described as some sort of numerical difference. That number only means anything in very special cases, and even then it's a stretch.

John-Paul,

I've wondered about the Fig. 4-21 (b) situation myself. I think you're right that it's legal but not practical. The primary datum feature might fully constrain 4 degrees of freedom, or it might not. It depends on the actual size. To me, that's a shaky design practice.

My objection to Fig. 4-21 (b) is that it shows the secondary datum feature in full (i.e. 3 point) contact with its simulator. While this isn't wrong, it's just one of the many possible configurations. The secondary planar datum feature only needs to have one-point contact with its simulator, even if there is slop on the primary. The figure implies otherwise.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[fsincox]

Belanger,
I think basically what you are implying is true, it can go either way depending on the actual application's numbers, if it is not controlled. I feel the standard has always recognized this since '82. This is also, why I disagree with the contention that calling out squareness of a small pilot is fool's play as some are promoting. On the contrary, I believe it is engineering's job as a good designers to look at the actual numbers and insure that what we intend to happen is designed to happen weither it is easy to inspect or not. That is one of the points I have been trying to get to in some of my threads.
Datum shift is now called that to separate it from the bonus tolerance concept, in '82 it was all thougth of as bonus but it still had to function differently. Remember all features to the same identical framework must also shift togeather as if a single pattern (except for the second line of composites added/clarified in '94)

 

[fsincox]

dingy2,
"I am a bit confused about the use of MMB on a datum plane? One does not gain any bonus or shift on a datum plane even it it was qualified with a feature control frame using perpendicularity, as an example."
Do you have a specific example from the 2009 standard of the application of MMB on a datum plane, or standard text paragraph/section reference?

 

[fsincox]

dingy2,
Thanks, I think I found it 4.16.4 (figs 4-30 & 31)?

 

[axym]

Dave (and fsincox),

MMB could be applied to a secondary or tertiary planar datum feature that is located relative to a higher precedence datum. See Fig. 4-30 and section 4.16.4. The shift would be purely rotational in this case.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[dingy2]

Evan:

I do agree that MMB (new term for 2009 edition) can apply to plane that could be a secondary or tertiary datum which would affect the part's orientation.

Who said that there were no real changes in the 2009 editon relative to datums? I think the changes are massive and will confuse so many people. I don't know how many times that I read through that section but each time, one learns a bit more.

Dave D.

http://www.qmsi.ca

 

[Belanger]

Dave,

Reread my post. I said that "the stuff about a datum referenced at MMC being primary, secondary, or tertiary didn't change in 2009." Yes, there were many conceptual changes. But that wasn't one of them.


John-Paul Belanger
Certified Sr. GD&T Professional

http://www.gdtseminars.com

 

[NormCrawford]

Using an MMB (2009) or MMC (1994) modifier for a primary datum reference is very common and in reality often a good design practice if it allows the additional available functional shift (release of one or more degrees of freedom). It is a simple question to ask; "Does the primary datum feature of size have a clearance fit with it's mating feature?" If so, then MMB/MMC should be applied because that is what is going to happen on assembly. The datum at virtual condition rule with regard to primary, secondary, or tertiary is of course a great deal to get into. But for simplicity, the datum at virtual condition rule can only apply to a primary datum reference when DML or DMP straightness is applied to the datum feature (DMP flatness for 2009). With regard to secondary and tertiary datum features of size, nothing has physically changed there either. But one still has to figure out whether the applicable virtual condition is orientation or location (mentioned in an above post in very good practical terms)depending on the relationship to the common higher ordered datum feature reference. The bottom line is that 2009 did not change reality!
What it trys hard to do is provide symbols to state the control of datum features to be fixed or translate, which includes planar surfaces as datum features having an MMB or RMB modifier. It does that fairly well. I think the major confusion will be recognizing whether or not the planar datum feature (secondary/tertiary) is an "offset" planar datum feature or not. If not, datum shift is available. If it is offset, shift is not available. (figs: 4-30b, 4-31c)

Norm Crawford
GDTP-S
Applied Geometrics, Inc.

http://www.GDandT.com

 

[axym]

Dave,

Obviously there were a lot of changes in 2009 with regards to datum reference frames. All John-Paul and I have been trying to say is that the concept of datum SHIFT (the topic of the original post) did not change.

If there is clearance between the datum feature(s) and the datum feature simulator(s), then the part and the datum reference frame can shift relative to each other within that clearance. In all but the simplest and most ideal of cases, the available shift is a complex mixture of translation and rotation that is not representable by a numerical value. This is unchanged from '94.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca