Straightness tolerance - Effect on material boundary size

[Jacob Cheverie]

I have attached a very basic image of a part that I am dealing with. Datum feature C is a cylinder, datum feature D is a cylinder, and datum feature E is a cylinder. These three cylinders are mutually orthogonal.

(1) Datum feature D has a Straightness tolerance at MMC.
(2) Datum feature C has a Position tolerance at MMC with respect to datum feature D at MMB.
(3) Datum feature E has a Position tolerance at MMC with respect to primary datum feature C at MMB and secondary datum feature D at MMB.

I am trying to establish the boundary size of D at MMB for the third geometric tolerance as listed above and I am running into an issue. The boundary size of datum feature D would be its MMC size plus any applicable geometric tolerances to higher precedence datums (per Y14.5-2009 4.11.6/2018 7.11.6). In this case, the designer does not include a geometric tolerance on datum feature D with respect to datum feature C. They actually have it the other way as is mentioned in case (2).

My question is as follows: To determine the MMB size of datum feature D, am I allowed to use the Straightness tolerance to increase the boundary size even though the Straightness tolerance is not defined to be in effect while constrained to datum feature C at MMB? Would I be violating datum feature precedence by adding this tolerance to the boundary size?

Thank you.

 

[Jacob Cheverie]

Here is the image.

 

[greenimi]

To determine the MMB size of datum feature D, am I allowed to use the Straightness tolerance to increase the boundary size even though the Straightness tolerance is not defined to be in effect while constrained to datum feature C at MMB? Would I be violating datum feature precedence by adding this tolerance to the boundary size?

If I understand your question(s) correctly I would say: No, because you need the size of the related actual mating envelope and you (currently) you don't have such a relationship/ related.
If I am wrong I will stand corrected.

 

[Belanger]

This looks very much like a part I've dealt with in the past. I would say that the straightness tolerance does still feed into the MMB for the position tolerance of the left-hand hole. That's because straightness doesn't relate that datum feature to anything else -- it's simply a form tolerance -- so it's not part of the "higher-order" discussion.

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

 

[greenimi]

Jacob,

Why position is shown on datum feature C definition?
IMHO should be perpendicularity. Am I correct?

 

[Jacob Cheverie]

Greenimi,

The Position callout on C with respect to D will control Perpendicularity as well as any error in location. The axis of C should intersect the axis of D, but may not. Imagine if the hole was drilled off the axis (location) and at an angle (perpendicularity).

 

[pmarc]

Jacob,
Do you know what is the reason that datum feature C is controlled relative to datum feature D first (i.e. C is of lower order of precedence than D), but then C is used as higher order of precedence datum feature than D in the position callout for E? Couldn't the order be reversed in the position callout for E to D(M)|C(M)? This would make things much simpler.

I am asking because in my opinion this would make all the MMB calculations much simpler. As it is now, I am afraid it is not the straightness tolerance that you should be thinking and worried about. Calculation of real MMB size of D in the |C(M)|D(M)| DRF requires relatively complex geometrical analysis to account for scenarios where C has been produced tilted relative to D within the its position tolerance. Also, this would have to include the length of datum feature D in the consideration.

 

[Jacob Cheverie]

Pmarc,

Everything else on the print is treating D as primary datum. This is an absolute blunder in my opinion as it is called out. We are seeing major errors in inspection due to the tilt of C as you are getting at. If I am not mistaken, I’ve used basic trigonometry to show that it’ll only take about 6’ of a degree of tilt on C relative to D to throw the Position of E OOT.

The problem is, the customer refuses to entertain the idea of changing the print. So we are stuck with what we have. I am being asked to find a way around it instead of actually solving the poor design. If I measure to the print, we see a lot of nonconforming parts. Using the Straightness tolerance on D to enlarge the MMB seems to help a bit, I just can’t decipher whether it is valid.

 

[Belanger]

I suspect that the reason for the confusing sequence of datum references is that it was a mish-mash of weighing the function of the part vs. the inspection of the part. Datum feature D is probably not all that functional -- it spaces apart the two holes but it probably doesn't touch anything (please correct me if needed, Jacob). That's why datum C was bumped up to the primary spot for the left-hand hole.

But along the way, someone was trying to figure out how to tolerance/inspect the right-hand hole, and they thought that it would be best to grab onto the main bar first and then look for the RH hole.
The result of those two ideas is the confusing datum structures.

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

 

[Jacob Cheverie]

Belanger,

You hit the nail on the head. The datum feature D is not functional in it's application. It makes sense why they may have thought to treat C as primary, but it is opening up a lot of room for error if the parts are inspected following Y14.5 practices as the print is drawn to.

 

[greenimi]

The Position callout on C with respect to D will control Perpendicularity as well as any error in location

Jacob,

I would still argue that the relationship between C and D is orientation and not location. I would even say that you are using wrong callout because you cannot tell be how far away C should be from D (there is no basic dimension between them). There is only 90 degrees basic angle, hence the correct grammar should be perpendicularity. Again, I am talking about the right hole (not the left one).

My argument stand still regardless if you change (reverse the order) as per pmarc's suggestion below.

Couldn't the order be reversed in the position callout for E to D(M)|C(M)?

The problem is, the customer refuses to entertain the idea of changing the print.

Well, that is a very common scenario. Welcome to the club.

 

[Belanger]

greenimi -- If the axis of that RH hole intersected with a datum that is a plane, then perpendicularity would have been the symbol to use. But the OP has datum D as an axis. And there is a basic dimension for the location of datum feature C to datum D: it's an implied zero.

To illustrate, simply hold your index finger from your right hand pointing toward your screen to simulate the hole's axis. Hold your left hand's index finger sideways, parallel to your screen (in the direction of datum axis D). Now, move your right hand up or down -- notice that it's still perpendicular to the datum, but its location could be off. Thus, the position symbol is correct if they wish to control orientation and location (however, based on function the hole should probably be the datum feature in this case, but that's a different issue).

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

 

[3DDave]

It seems unlikely to me that there is any benefit to using MMC/MMB as a modifier on any of the FCFs.

That aside, it seems so badly done as to have no meaningful resolution. It would be more meaningful to separate the location tolerance from hole to hole from the location tolerance of each hole relative to the nominal rod. These are independent considerations and tying them the way they are depicted makes unrealistic requirements.

In intepretation - there is a location relationship between the right hand hole and the nominal rod; anyone who has tried to drill a hole centered on a dowel will understand the frustration that can occur without careful preparation, no matter how perpendicular the resulting holes and their axes are to one another.

The feature chosen as datum feature D is functional, at least to the point that were it to be tangential to the two holes the part would likely not function in it's next assembly. There must be some limitation for mismatch in offset between the mutual pairs of axes.

 

[greenimi]

If the axis of that RH hole intersected with a datum that is a plane, then perpendicularity would have been the symbol to use. But the OP has datum D as an axis. And there is a basic dimension for the location of datum feature C to datum D: it's an implied zero.

Ok. I understood. Thank you very much.
For me, at least initially, looked like half orientation (along axis of D) - half location (normal to the axis of D)

Even in J-P plane example,

If the axis of that RH hole intersected with a datum that is a plane, then perpendicularity would have been the symbol to use.

I think if the plane is perpendicular to the axis of RH then as J-P said, perpendicularity is the symbol to be used, but if the plane is nominally parallel to the axis of the hole (like rotated 90 degrees, then, probably still position is the "correct" symbol)

Pmarc,

May I ask you: is "my misunderstandings" one of the reasons on why ISO introduced orientation plane indicator?
Without hijacking the thread, I am thinking about fig 84 (ISO1101:2017) where are two positions shown (one with Ø symbol and one without) both of them to D (coincidence to the OP's primary datum feature) and then two orientation plane indicators to the same datum D.

 

[pmarc]

greenimi,

I don't think it was the reason why ISO introduced orientation plane indicator. The reason was rather to not have to rely on purely pictorial representation of the orientation of the tolerance zone, but make it automatically downstream consumable by machines like CMM etc.

So if you take Fig. 7-28 in Y14.5-2009 as an example, the two position tolerances cannot be seamlessly consumed by CMMs because the direction of the dimension and extension lines means nothing to CMMs.

 

[Belanger]

if the plane is nominally parallel to the axis of the hole (like rotated 90 degrees), then, probably still position is the "correct" symbol

I agree, because if the hole and the plane are running in the same direction, then there is a distance between them, and that's position's main job.

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

 

[Jacob Cheverie]

Digging in a little deeper to this issue, I came across Y14.5-2018 7.9 (pg. 72) which states

...
To make it possible to calculate the true geometric counterpart boundaries of each datum feature in a datum reference frame, a relationship between the datum features shall be specified. Tolerances applied that affect a datum feature or the relationship between datum features include the following:
(a) primary datum feature(s) size and form (see Figures 7-2 and 7-4) and/or the location between features in a pattern used to establish the primary datum (see Figures 7-16 and 7-17).
(b) secondary datum feature(s) size, orientation, and/ or location, as applicable, to a higher-precedence datum (see Figures 7-2, 7-4, 7-18, 7-32, and 7-33); where a pattern of features serves to establish the secondary datum, the location between features in the pattern and the pattern’s orientation, location, or both to a higher-precedence datum shall be applicable.
...

(b) seems to imply that the form of a secondary datum cannot be used to calculate the true geometric counterpart boundary. Any thoughts?

 

[greenimi]

seems to imply that the form of a secondary datum cannot be used to calculate the true geometric counterpart boundary. Any thoughts?

I think if secondary is modified at MMB/ LMB cannot, but if left RMB it can. ( fig. 7-24/2018). My opinion.

 

[SeasonLee]

I think if secondary is modified at MMB/ LMB cannot, but if left RMB it can. ( fig. 7-24/2018).

Why only RMB will calculate the form tolerance? I'm trying to learn here, hope it will not hijack the thread.

Season

 

[greenimi]

Season,

I'm trying to learn here,

Me too.

See below.

https://www.eng-tips.com/viewthread.cfm?qid=465740

axym (Industrial)
6 Mar 20 16:59
Hi All,

This is what we sometimes refer to as a "swamp".

Y14.5 indirectly weighed in on this question in the 2018 standard. There is an updated version of what was Fig. 4-16 in 2009, with example calculations for MMB. The new version in 2018 (Fig. 7-22) adds a straightness tolerance at MMC to feature D, but this tolerance has no effect on the MMB's. This is not mentioned in 7.11.6.1 on Determining the Appropriate MMB, but it is clarified in 8.4.1.3 on Derived Median Line Straightness:

"When the straightness tolerance at MMC is used in conjunction with an orientation or position tolerance at MMC, the specified straightness tolerance value shall not be greater than the specified orientation or position tolerance value and does not contribute to the IB or OB of the position or orientation tolerance. The collective effect of the MMC size and form tolerance produces a VC, OB, or IB resulting from the form tolerance but does not affect the IB or OB created by any orientation or position tolerances on the feature. See Figure 7-22."

So we're not allowed to specify a straightness tolerance at MMC that exceeds the position tolerance at MMC. The boundary resulting from the straightness tolerance must somehow get absorbed by the boundary for the position tolerance. Does this mean that specifying Independency, which would allow even more straightness error, would not be allowed?

There are new examples of boundary calculations for LMC and RFS cases as well. In the RFS case (Fig. 7-24), the straightness tolerance on feature D actually does affect its boundary!

I have not been able to make sense of this.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca