Non GD&T dimensioned features & assumed tolerances 3

[Runz]

We have several drawings that contain partial GD&T positional Tolerance. Most of the diametrical dimensions do not have any GD&T positional tolerance. The titleblock says the drawings are dimensioned and toleranced per ASME Y14.5-1994. There are also tolerances given for 2 and 3 place decimals (.03 & .010) and angular dimensions (+/- 1/2°)
My question is, what are the none GD&T dimensions held to positionally? See pic attached. Are they assumed located to a true center axes? What diameter do is used to find the center axes? Depending on the diameter you use, could it not change your inspection results? Are there assumed positional locations on these diameters or only size tolerances? Just a little confused as to what is assumed when there is no GD&T on a dimension.
Thanks,

John

 

[pmarc]

Kenat,
You said:

Even if you ignore that or think around it and want to invoke the block tols, if they vary with # of decimals how do you know which one to choose?

This was a point of my first response within this thread. You do not know which general linear tolerance to choose because you do not know how many decimals implied 0 has. But that does not mean the control is not there. There is still general angular tolerance which may be used by somebody to control the coaxiality.

Don't get me wrong. I do really believe the print is poorly dimensioned. I just want to reveal how many different interpretations may hide behind this scheme, which - by the way - I see very very often. Too often, I should say.

As to 90[°] - paras. 1.4 (i) and (j) tell that 90[°] angle applies but do not say what defines the tolerance for this angle. So my question is still the same: what would limit mutual perpendicularity of those two features if there was nothing specified in the print area besides limits of size and only general angular tolerance defined in the title block?

 

[KENAT]

Since it is being expressed as an angle, it would seem appropriate to invoke the angular tol from the tol block. However, I'm sure someone out there will argue that it's more akin to angular GD&T so that linear tol should be applied and in that case which # of decimals...

Fact is the OP's drawing is incomplete per the invoked standard.

Posting guidelines faq731-376

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

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

 

[Belanger]

I would never think of an implied 90º angle as having anything to do with linear tolerances or a number of decimal places. Appealing to the general angular tolerance seems to cause no trouble, in my view at least.

Pmarc ... since you posed the question, now I'm curious about your own answer!

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

 

[pmarc]

J-P,
I asked the question to Kenat hoping he would say that for two features shown at right angle on a print the general angular tolerance controls their mutual orientation. Having this answered I would then say that in my opinion general angular tolerance could be applied to nominally coaxial features as well, because the only difference is that the features are shown at 0[°] to each other and not 90[°].

Personally I think general angular tolerance also causes troubles. It may be enough in some cases, but contains at least two shortcomings:
1. It is ambiguous since it is not known which feature to use as a base for angle measurement;
2. The tolerance zone is not fixed in width, which in fact can significantly affect functionality. The zone increases as it gets farther from the apex of the angle.

 

[axym]

Hi All,

I always thought that the general tolerances apply to the dimensions (linear or angular) that are actually shown on the drawing, and not to dimensions that are not shown. If the general tolerance depends on the number of decimal places shown, then it has to work this way.

If a general angular tolerance applied to undimensioned angles, then it would control the angle between any two features on the part. This seems impractical - you could pick any two features and say that the angle between them must be within +/- 1 deg or whatever the general tolerance is.

What if the general tolerance for linear dimensions didn't depend on the number of decimal places, and was just a constant amount like +/- .030" or something? Would this apply to dimensions that are not shown on the drawing, or only to the ones that are shown? Again, it seems impractical for this tolerance to apply to dimensions that are not shown - then the distance between any two features would be subject to it.

I know we're not talking about ISO, but how do the general tolerances in the ISO 2768 standard apply? Do they just control dimensions that are explicitly shown on the drawing, or unspecified dimensions as well?

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[KENAT]

pmarc, I actually considered that before replying and read 1.4 I & J carefully but they are very clear in talking about 90° not shown aligned or 180...

As to iso2768 based on the standard I'd think only dimensions explicitly shown (at least part 1). However, my first encounter with that standard involved folks that thought otherwise.

Posting guidelines faq731-376

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

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

 

[Belanger]

Evan and pmarc ... a general angular tolerance applied to undimensioned angles, ONLY in the special case of 90º angles, because of paragraph 1.4 in the standard.

Pmarc, I see coaxial relationships as linear, so the general angular tolerance doesn't apply. General linear tolerances could apply, if we didn't have the issue of decimal places. The angular tolerance (usually) doesn't deal with decimal places.

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

 

[CheckerHater]

To axym:

ISO 2768 Part1 is dealing with non-toleranced dimensions, Part2 is dealing with non-specified geometric relations, including circular Run-out. Coaxiality is not specified.

I like to refer to this picture:

http://www.globalspec.com/reference/43769/203279/chapter-16-general-geometrical-tolerances

It shows "invisible" GD&T implied by invoking 2768 very nicely.

 

[pmarc]

I love the example CH linked to. My 2 favorites are:
1. Hole controlled by position callout suddenly interpreted as a datum feature C per main view of bottom sketch;
2. Perpendicularity callout 0.2 wrt C shown on the right view of bottom sketch. How somebody can even suppose that this tolerance must be considered relative to C and not to A for instance?

 

[powerhound]

Is it okay to specify parallelism without a datum reference in ISO?

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[CheckerHater]

There is no view showing flats being perpendicular to [A], so there is nothing to imply.

On the other hand, THERE IS view showing flats being perpendicular to axis of the hole, so perpendicularity is implied.

Good enough for me.

 

[fsincox]

I am with CH on this, say: +/-2 degree on angles, is just as bad, which is the datum for that?? General tolerances represent a compromise where ever they are used, to then turn around and say they are not perfect is redundancy in itself.
Frank

 

[pmarc]

CH, Frank,
I see your point, however do not buy it. Would you really interpret it in this way if there were no dashed (general) tolerances shown? Look at a flat surface at the very left of the part. The interpretation is that circular runout 0.1 controls its relationship to datum feature B. My question is could not it be parallelism to A instead since both features are shown in the same view too?

powerhound,
Parallelism callout on the left view has datum reference. Notice that black triangle is attached to one of two faces which means this face is considered to be a datum feature. This way of FCF presentation was legal up till 2004 when new revision of ISO 1101 was issued. Since then it has been withdrawn. Unfortunately the picture we are talking about right now comes from ISO 2768-2:1989, so way before the method was clearly forbidden.

 

[DeanD3W]

Hey,
Nobody pointed out that there is a missing diameter symbol in the one position tolerance that does appear in the drawing the Runz referenced in his original post.

Just add a few more position tolerances and all would be well.

All good comments regarding how unwell we are if the ambiguity of "default tolerances" are allowed to exist.

Dean

http://www.d3w-engineering.com

 

[powerhound]

I was talking about the parallel callout on the view on the right. Pretty much the end view of the part.

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[powerhound]

Never mind. I see what you're saying about the sucker foot. Is this legal per ISO?


Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[pmarc]

powerhound,
Yes, I was also thinking about parallelism callout on the right view. Do not why I said it was on the left view.

 

[pmarc]

Like I said, it has been illegal since 2004.

 

[powerhound]

Never mind again. You already answered my question...sheesh. I need to slow down a little today.


Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[powerhound]

Sorry to resurrect a dead thread. I didn't see where this was brought up but while 1.4(i) says that angles depicted at 90 degrees should be implied 90 degrees. It doesn't say what tolerance to use. Paragraph 2.1.1.2 clears this up by saying that the tolerance on implied 90 degree angles comes from the default tolerance block.

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X5
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II