Examples of Equivalent Dimensioning & Tolerancing Schemes

[pmarc]

Hi,

Throughout the years I have come to a conclusion that in general there are not many examples where changing dimensioning and tolerancing scheme from one to another would keep the geometric requirements for the system unchanged.

One example where this conclusion would not be true is changing from perpendicularity wrt A to total runout wrt A when applied to a flat face normal to datum axis A.

Another one would be a simple bushing where its ID and OD are controlled with the same +/- tolerance, and then it does not really matter which of the features will be datum feature A and which will be controlled with position or runout relative to A.

I have some more, but I would like to see what others can offer. So could anyone share some examples?

 

[pmarc]

Feel free to call the translation modifier whatever you like.

I just tried to explain that from Y14.5 perspective...:
1. CRDF concept does not override default basic locational relationship between datum feature simulators, and the translation modifier does.
2. Your scheme with CDRF is not equivalent to the scheme shown in fig. 4-19, but is identical with the scheme shown in fig. 4-9.

 

[CheckerHater]

I have to apologize to 3DDave but I cannot avoid using word "demagoguery"
If datum C on Fig. 4-9 only constrains 1 degree of freedom, then there is no need to translate.
Pins B and C on 4-9 will fight each other.
Pin C on 4-19 is given extra DOF.
It may be of purely academic interest to some, but I was designing fixture for condition like that: round pin going into round hole, but pin itself placed on slide, so it will not fight other datum simulators. It's all about degrees of freedom.

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

 

[Belanger]

Pmarc, yes that's correct. I was just getting nitpicky because this thread is about equivalent callouts, and CF on a surface -- without a flatness tolerance -- is meaningless.
It's like those goofy notes that say "all diameters shown in-line must be coaxial."

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

 

[3DDave]

Figure 4-9 allows a pin at C to start at diameter = 0 and then expand from there until it makes contact. Just like a real expanding RFS pin it may require the part to deform, but the standard places no limits on the amount of force expansion can exert or the degree of force against which to stabilize.

In total, the expanding datum feature simulator for C only constrains 1 DOF in [A|B|C]

I see no practical use in such a scheme, or the one explained in Figure 4-19, but I'm not trying to sell the translation modifier.

In real parts, using a pair of interference fit pins, the result will depend on elastic or plastic deformation of the parts involved and if fine control is desired, the end placement will not be predictable from an 'expanding pin' evaluation.

 

[CheckerHater]

Round pin placed in a round hole constrains 4 degrees of freedom - always.
Just nail 2 pieces of wood together.
But since the game of the day is to deny the obvious, I am out of here.

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

 

[pmarc]

CH,
I recommend following experiment:

Step 1:
Take 2 flat pieces of the wood and put them together. Hold one of the pieces still and try to play with the other piece.
Question: What kind of movement will be possible?
Answer: The second piece will be able to freely translate in x and y directions and freely rotate about z axis (normal to the plane of contact). This means that 3 degrees of freedom have been constrained and there are 3 degrees of freedom yet to constrain.

Step 2:
Nail the pieces together using one nail. Hold one of the pieces still and try to play with the other piece.
Question: What kind of movement will be possible?
Answer: The second piece will be able to rotate about the axis of the nail. This means there is still 1 rotational degree of freedom to constrain and 5 have been already constrained.

Step 3:
Nail the pieces with second nail and holding one piece still try to play with the other piece.
Question: What kind of movement will be possible?
Answer: The second piece is now not able to move at all. This means the remaining rotational degree of freedom has been constrained by the second nail.

The second nail used in Step 3 of the experiment is working similar to how the datum feature C simulator in fig. 4-19 works. The nail did not constrain 4 degrees of freedom, it just constrained 1 because other 5 have been already constrained by putting two pieces together (3 DOFs) and nailing them with first nail (2 DOFs).

 

[3DDave]

pmarc - the place that analogy goes apart is this - the example nails are forming nominally match-formed holes in the pieces of wood and removing either nail adds one degree of freedom, even if it fixed two when installed. In essence the pair of nails is required to fix planar rotation, so they arithmetically each contribute half, but only as a pair.

Looking at the same experiment with predrilled holes, where the separating distance is not the same for the two pieces and the diameters of the matching holes in the two boards is not the same, placing the second nail results in a shoving match, where the relative location of the boards depends on which hole has a tighter fit on its respective nail, which nail is larger, how elastic each nail material is and how elastic the material the boards are made of is.

The problem with the Y14.5 rigid-body assumption is that using two RFS connections mechanically overconstrains the part and cannot be resolved by simply measuring the geometry. Making the assumption that overconstraint doesn't matter just to make it easier to check on a CMM does not resolve that problem. The similar diagram in 1994 used (M) such that selecting a fastener that would not mechanically overconstrain the assembly was possible; the shift to RFS in the 2009 version mechanically makes no sense.

The way they attempted to get around this is to use fictional expanding pins that will be just the right size to not apply any loads. I have seen pins that can work like that in match drilled holes that are then match reamed, but they suffered locating precision problems when elements along their length were allowed to expand separately in the mating parts. A full split sleeve was used to force a more uniform expansion, but that only works if the holes are precision machined post-assembly and match in size and location withing a few tenths of thousandths of an inch.

 

[Sem_D220]

If I may add my 2 cents here:

1. 2 holes as RMB datum features do not necessarily overconstrain the real application assembly, but they may overconstrain inspection assembly in a fixed gage. This is where either TM or CDRF might be useful.

2. TM may be useful in cases where the application assembly connects 2 different rigid parts, or in other words when the distance between the pins that fit into the holes is adjustable.

3. CDRF might be useful where there is a single mating part, which is non-rigid, contains 2 pins that should fit the holes without significant clearance, and it is intended to fit one pin before the other. Hence the specific degrees of freedom constrained by each hole.

Edited for more accurate phrasing

semiond

 

[Sem_D220]

In other words, your proposed scheme is not overriding any degrees of freedom and is no different from what is shown in fig. 4-9 in Y14.5-2009 - there is no need to specify [w] after letter C in both position callouts (just like there is no need to specify any degrees of freedom after A and B), because that is the only degree of freedom that can be constrained by datum feature simulator C. One might say your scheme is redundant in this case.

I disagree with that.
By the same logic, the scheme shown in fig. 4-45 is redundant too. But it isn't.

 

[pmarc]

I disagree with that.
By the same logic, the scheme shown in fig. 4-45 is redundant too. But it isn't.

In fig. 4-45 there is no redundancy because by default a cone as primary datum feature constrains 5 degrees of freedom. Fig. 4-45 says that is should constrain only 4.

In CH's proposed scheme not a single degree of freedom has been overriden. The scheme just lists degrees of freedom that will be constrained by each datum feature simulator without taking any override action. So it is like using fig. 4-2 and instead of leaving the position tolerance for the two holes as is, changing it to: |POS|Ø0.2(M)|D[z,u,v]|E[x,y]|F[w]|. Can be done but is redundant.

 

[pmarc]

Ugh! There should be |POS|Ø0.2(M)|D[z,u,v]|E[x,w]|F[y]|. 

 

[Sem_D220]

Agree about fig. 4-45. It was my oversight.

I still think that CH's proposed scheme may be useful, because without the CDRF, in terms of which datum constrains which DOF in the inspection assembly, the dimensioning scheme in fig. 4-9 might be eventually (after assembly) same as if it was with |A|B-C| datum DRF (another equivalent scheme to the list?). It is because you can arrange the datum precedence order as you wish in the drawing, but eventually physics will dictate behaviour according to it's own rules. If CDRF doesn't allow non-basic location of datum feature simulators, then CDRF or not, I doubt the ambiguity can be prevented in this particular case as long as both holes are called out RMB, and even if the tetriary datum was called out MMB, there can still be an as-produced hole sizes combination within tolerance where the problem is not prevented. But at least it tells the inspector to try maybe in other cases that scheme could prevent ambiguity?

I'd like to emphasize that the ambiguity I mentioned can occur in the inspection assembly, but not necessarily in the real application assembly.

 

[pmarc]

I disagree that |A|B|C| is the same as |A|B-C|.

And with regard to: "It is because you can arrange the datum precedence order as you wish in the drawing, but eventually physics will dictate behaviour according to it's own rules". That is the whole trick with proper datum feature selection - they should not be selected in the order that one just wishes them on the drawing, they should be selected in the order that mimics the physically reality as close as possible.

 

[Sem_D220]

"they should not be selected in the order that one just wishes them on the drawing, they should be selected in the order that mimics the physically reality as close as possible"

That is always true. And my point - once the datum features and their precedence selected according to the physical reality, one can use methods such as CDRF to communicate to the manufacturer and inspector that they have to make sure that during verification of drawing specs, the part will behave according to the physical reality in real assembly. The fixture will have to be made accordingly, otherwise ambiguity might occur.

For fig 4-9, what makes you sure that the A|B-C| scenario can be avoided by those who act according to the A|B|C specification? (Hope my question makes sense)

 

[pmarc]

For fig 4-9, what makes you sure that the A|B-C| scenario can be avoided by those who act according to the A|B|C specification? (Hope my question makes sense)

If someone trully acts (sets up part for inspection) according to the |A|B|C| specification, the |A|B-C| scenario will avoid itself.

I agree that CRDF is to communicate to the manufacturer and inspector that special attention is required during verification of the part, but if the special attention is in fact not needed, like in CH's proposed scheme where CRDF does not change anything compared to fig. 4-9, is there really a need to decorate the drawing with symbology that adds no value?

Unless we are saying that the manufacturer and/or inspector may have difficulties to intepret the meaning of regular |A|B|C| DRF specification, and therefore the redundant CRDF is to help interpret the DRF properly. But in this case we might as well say that in order to communicate that for example Rule #1 applies to all holes in fig. 4-9, zero @ MMC DML straightness callout should be added to each hole.

 

[Sem_D220]

I may be wrong, but I don't see how the A|B-C| scenario can always be avoided by following the specifications in the drawing. Yes, the inspector will assemble the datum features to their respective simulators according to the correct order: first B, making maximum possible contact with the expanding pin, and making sure translations are constrained, then C, also expanding till maximum contact, and yet hoping it only does "clocking". (EDIT: In case someone will decide to nitpick - I know A is the first in order, above I meant what happens after A is already in contact). But once the assembly is done, if datum feature hole C happened to be produced at same or smaller size than datum feature hole B, then not only the B-C scenario won't be avoided, B and C may in fact switch roles. Nothing in the inspection assembly can prevent hole C from taking perpendicular loads and stopping translations.
That is while for the real application, all kinds of means are available to secure the right function.

 

[Sem_D220]

There is a small inaccuracy in my last post. The problematic situation will occur not when hole C is produced at same size or smaller than B, but when the fit achieved between hole C and it's expanding pin simulator will happen to be as tight or tighter than the fit acheived between hole B and it's expanding pin simulator. I don't know how it can technically be avoided.

 

[pmarc]

If in a real assembly there is a chance that datum features B and C will switch roles, this means that |A|B|C| should not be defined on the drawing in the first place.

 

[Sem_D220]

In the real assembly all kind of things may or may not happen, including a the |A|B|C| function fully guaranteed (preferably). That depends on both the design of the part and on the design of the mating part(s). The drawing doesn't specify exactly how the part should function, but it must specify how it's degrees of freedom should be constrained in gaging assembly. Since the fixture and the mating part(s) are different entities, what is guaranteed in real assembly by means of mating part(s) design, may be be guaranteed for the fixture by means of CDRF specification.

 

[pmarc]

The drawing doesn't specify exactly how the part should function, but it must specify how it's degrees of freedom should be constrained in gaging assembly.

The drawing should exactly represent function of the part and gage design should reflect that as close as possible.

If datum features, their precedence and material boundaries at which they are referenced on the drawing do not reflect how the part is oriented and located in real assembly, then no matter how precise the gage is and how well it simulates specified datums, it will not prevent from accepting non-functional workpieces.