RFS and Projected tolerance zone concept on threaded holes 1

[bxbzq]

Hi,

I'm now working on some drawings where threaded holes need to be toleranced. I'd like to use projected tolerance zone concept in the positional tolerance specifications. I looked over the '09 standard and some examples in my training materials, all threaded holes using projected tolerance zone use together with MMC concept. The idea of the fixed fastener formula is that the the threaded hole and clearance hole share the same VC which in formula equals screw OD. On the other hand, some time ago I learned from this forum that using RFS concept to specify positional tolerance for a threaded hole makes more sense because the threaded hole has self centering function. I think I can still use RFS together with projected tolerance zone and assign the tolerances calculated from the formula to make the threaded holes' IB equal to screw OD. But, is this good practice?
I'd like to know what others think.

 

[CheckerHater]

I was looking for years for solid argument for or against using RFS with threads. There is always some slack left for the opposite point of view.

Here is a little bit of insight - not the the last instance of truth though:

http://files.engineering.com/getfil...c8f1e7d02e9&file=Projected_Tolerance_Zone.pdf

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

 

[greenimi]

From Mark Foster •
"Threaded features are deceptively complicated little creatures, and the standard somewhat "glosses over" how they are to be treated. The standard does say that, for screw threads, the default "feature" to which a geometric tolerance applies, or to which the datum feature symbol is referring, is the Pitch Diameter. That information is somewhat helpful in that we now know that the default is not the major or minor diameter, but none of these diameters, major, minor or pitch, is truly a diameter at all. They are really helixes. The "diameter" is just the imaginary cylinder that would capture said helix. So it is next to impossible (not impossible, just close) to actually measure the actual size of the "actual" pitch diameter cylinder.

There is one camp of folks who believes that all screw threads should be referenced RFS/RMB since one, screw threads have a tendency to center up on each other's pitch diameter axis when threaded together, and two, since you cannot really easily measure the actual size of the actual PD. While those are valid concerns/arguments, I do not fall into that camp.

Essentially, the "other" camp (the one I lean toward) says that straight threads (i.e. not tapered -- "pipe" -- threads) should be referenced with the MMC/MMB because there IS clearance between the threads as they are threaded together, and while they do have a tendency to want to center up on each other, you can -- without cross-threading -- screw two threads together where their PD's are not perfectly coaxial. And you can do that even more when the actual threads are produced with their PD's further departed from their respective MMC/MMB sizes.

Now, having said all of that, I do NOT believe that we should be attempting to measure the actual mating sizes of PD's and making use of the "bonus" tolerance in the same fashion as if they were clearance holes. In other words, the "bonus" tolerance for threads cannot be readily measured (and therefore calculated). The bonus tolerance for screw threads really manifests itself in gages. That is, if we were to build a functional gage to check the location of a threaded feature, we would not want to have "expanding threads" on our threaded gage elements (as in RFS/RMB) since that would not be the functional interface anyway. Use of MMC/MMB in the specification of those threads allows me to use straight threaded gage elements. You "just have to know" that the MMC/MMB specification does not apply like a linear calculation the way it does for clearance hole applications. Mark"

Threaded holes derive a negligible, non-quantifiable amount of datum shift from their pitch diameter. The theoretical max datum shift is the difference between PD max and PD min, but in reality is much smaller than this due to the above self centering nature. So, the screw/shaft inserted into a threaded hole has ability to wobble inside, but the amount of this "wobble" is NOT the difference between PD max and PD min.

 

[greenimi]

And one more from Marik Foster:
"I respectfully, and only partially, disagree with your opening statement regarding threaded features self centering. Yes, I agree that threaded things have a tendency to self-center as you tighten them together, but I disagree that this action "...negates any additional location tolerance...." and that "...there are no bonus tolerances gained whatsoever." It is possible, and even probable in most assemblies, that the two pitch diameters -- that of the threaded hole and that of the threaded bolt going into said hole -- are *not* completely coincident once assembled together. Yes, they do have a *tendency* to center on each other, but they never quite make it exactly to completely be exactly the same. And, the more difference there is between those two mating parts' relative PD "sizes," the more mismatched their respective PD axes can be and still assemble together just fine -- without cross-threading. It's just that this relationship is not readily "measureable" since it is impossible to "measure" the "actual size" of a PD. Thus, the relationship is not a 1:1 relationship like it is when putting non-threaded features together, and thus, it is impossible to calculate the actual amount of "bonus" tolerance available.

It *is* however, possible to gage such requirements, and forcing inspection to make use of expanding gage elements when that is now how the parts will ultimately fit together is, IMHO, artificially tightening the tolerance specification from what it functionally needs to be.
"

 

[greenimi]

And the last one:

"Regarding MMC use on threaded features, let me put it this way: If you are only measuring, and never gaging, your orientation/location of your threaded features, then you could always treat all threads as RFS and you will probably not have any issues (except for perhaps your ability to measure consistently -- since that is another issue altogether). The reason is that the amount of total potential bonus tolerance is generally pretty small compared to the orientation/location tolerances, and any physical bonus tolerance that might exist is not measureable (only gage-able) and is not a 1:1 ratio anyway.

So if you went the rest of your life ignoring any potential bonus tolerance on any threaded features (and just a reminder that I am only talking about Pitch Diameters, here), then you will be conservative and probably not have any problems.

As I have mentioned in this thread and previously,there are two camps of people, those who believe that ALL threads should only be specified RFS (like you), and those who believe that only tapered threads should be specified RFS (like me). I think that we can coexist. As I have also mentioned previously, the bonus tolerance that physically exists between straight-threaded features' Pitch Diamters is not a really measure-able thing, and so really only can be assessed through gaging and actual parts, not through CMM measurement/assessment. "

 

[CheckerHater]

So, one camp believes that threads are to be specified RFS because bonus doesn't exist.
The other camp believes that threads are to be specified MMC because bonus is impossible to measure.

I'd rather join the camp that relies on metrology - if it's smaller than measurement error, it doesn't exist for all practical purposes.

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

 

[drawoh]

I was looking for years for solid argument for or against using RFS with threads. There is always some slack left for the opposite point of view.

Threads are self centreing. MMC/B and LMC/B have no meaning.

Projected tolerance zone is meaningful, and probably a good idea if you are clamping something thick to the tapped hole. The clearance hole must clear the fastener diameter, plus the tolerance zone. This zone needs to be projected if the fastener is long enough to clamp the thick part.

--
JHG

 

[KENAT]

The way I look at it is while MMC has limited application on threads - essentially just for inspection via hard gaging - it doesn't cause any real harm putting it on there even if the inspector doesn't end up taking advantage of it.

So I throw it on.

I'm sure someone will be all to willing to point out why this is horribly wrong.

Posting guidelines faq731-376

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

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

 

[3DDave]

At best threads are both - clearance applies while the fastener is being assembled and they are self centering when tightened. They are a hybrid. If the important case is whether the thread can assemble, then MMC; if the thread is guiding a feature that will also be used RFS, then use RFS.

I wonder about controls where the thread is allowed to cut into the mating part just a little, a common assembly condition that is a minority condition, but still encountered enough to not be rare. I find that it's worth .010 to .020 on the diameter for mislocated holes.

I think the best answer is to use MMC for feature acceptance to justify using fixed form simulators and pretend there is no clearance for location tolerance calculation purposes.

 

[bxbzq]

Here is another angle to look at the theaded hole case. If it is MMC like clearance holes in floating fastener case, you assemble and disassemble them several times, you will end up with several different bolt to hole inter positions. Threaded holes in fixed fastener case, however, you assemble and disassemble them several times, I tend to say you would get almost the same screw to hole inter position. But I don't have any data to support. Has anyone done anything similar?

 

[greenimi]

If it is MMC like clearance holes in floating fastener case, you assemble and disassemble them several times, you will end up with several different bolt to hole inter positions

Isn’t this true for every measurement? All measurements made for any reason using any technique at all are subject to uncertainties. Some estimating techniques are better than others. It's just that some techniques have more uncertainty built in than others.

Lets say you measure a functional plane
*THE* truly functional plane is probably somewhere in between the LSQ plane and the 3-highest-point plane in many applications (very rigid parts). Also, * truly* functional plane is really not just one plane, either. It would actually change from mating part to mating part, and even with the same two mating parts put together multiple times by the same person.

Also from Mark Foster (Applied Geometrics)
"As I said, *any* measurement we do is an estimate of the real thing. We will *never* be able to truly know whether or not we found the truly functional plane. We can only hope to come close to the truly functional plane, and hopefully in a repeatable and standardized way. Y14.5 says that that way is the 3-highest-point plane, but that is a theory in and of itself since we can never be sure that we have found that plane with our measurement techniques. But we can do our best to attempt to simulate that plane -- i.e. when using a CMM, which algorithm does the best job of simulating what we really should be measuring from .

And my previous ramble essentially ended with, be aware of what your algorithms do, be aware of what your goal is (i.e. what you are attempting to simulate), do the best that you can (and document it), and properly account for whatever uncertainty you may have created with your chosen method."

 

[3DDave]

bxbzq, When I assembly a bolt into a threaded hole the bolt rattles around a lot. That's the assembled condition. When I tighten the bolt it aligns to the thread except where it is leaning against the side of the hole it passes through, so that's the tensioned condition. If you need better control than that, don't use a bolt.

Perhaps if you have a case where you think it matters then some calculations can be applied to see just how much it matters.

 

[bxbzq]

greennimi,
I'm afraid I don't get your point. Let's say my part has ø7 ±0.1mm clearance hole and a M6 bolt goes through it. In extreme conditions, the bolt-to-hole inter position can go 0.1mm misalignment. But if I measure an as-assembled piece and get 0.1mm measurement variation, my measurement does not pass gage R&R, or any method of MSA.
I guess we are talking about different variation levels. Measurement error is another animal.

 

[greenimi]

Let's say my part has ø7 ±0.1mm clearance hole and a M6 bolt goes through it. In extreme conditions, the bolt-to-hole inter position can go 0.1mm misalignment. But if I measure an as-assembled piece and get 0.1mm measurement variation, my measurement does not pass gage R&R, or any method of MSA.

No comment. I agree.