Tolerance on implied 90 angles.

[PeterStock]

I have a question that is derived from a drawing I recently checked. There were two values in the title block for angles, one of +/-2 degrees if the angle was stated as a whole number and one of +/- .5 degrees if the angle showed a decimal. My question is which applies in the case of an implied 90 degree angle?

My thought is that if I was making the part I would assume +/- .5 degress, but if I was inspecting the part, I would use the +/- 2 degrees as my accept/fail criteria. I am thinking that the 2 degrees would be what it would be rulled as in court (if it came to that).

What do you think?

Peter Stockhausen
Senior Design Analyst (Checker)
Infotech Aerospace Services

http://www.infotechpr.net

 

[Belanger]

The standard says that all undimensioned angles that appear as right angles are to be interpreted as 90º.

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

 

[flash3780]

It's my understanding that the implied 90deg angle is basic, so has no tolerance directly. Datums, however, are exactly perpendicular to one-another.
So, if you're thinking of a simple cube, you'll have to apply datums to three sides to fully define it. The maximum angular error would be dictated by the size tolerance (i.e. the angle created when one corner is at the maximum size and the other is at the minimum size).
[tt]
/--------/
/ /|
/--------/ |
|<--XX±X-->|
[/tt]

 

[flash3780]

[tt]^
|
|
|[/tt]
That's Rule #1, right? Perfect form at MMC.

 

[Belanger]

No Flash -- in fact Rule #1 doesn't help in that situation. Rule #1 says that form is controlled by size (or as you phrased it: perfect form at MMC). That's great if we're concerned about flatness. But when we say size, that is simply on an individual basis. So the width, length, and height of your parallelogram could be at their max sizes but that says nothing about the angle on the corners!

Think about it: if you get out a measuring tape and just measure the height, that may be at MMC but that's OK. If you then use the measuring tape to check the width, it too could be at MMC and that's OK. But the angle could be 5º and you might never know it.

This is why there should always be a general note to control those implied 90º corners, since Rule #1 only controls form (flatness), not orientation (perpendicularity).

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

 

[CheckerHater]

To flash:
Implied 90 deg. angle is basic only if it's angle between datum features, otherwise general tolerance applies.

Also (to pour some gasoline into campfire), if you invoke ISO 2768 part 2, implied Perpendicularity will override general angle tolerance.

 

[Belanger]

Hater -- In ASME, implied 90 is basic only if there is a geometric tolerance imposed on that angle. Just calling two surfaces datum features doesn't tolerance them; the title block still applies.

Maybe this is where the ISO 2768-2 comes in? I'll have to look that up ... maybe that's what you are saying in your first statement.

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

 

[CheckerHater]

Sorry Belanger, I was responding to flash's idea of "simple cube" with datums applied to 3 sides, not making general statement.

Nevertheless your comment is very interesting. In your opinion, on the enclosed picture which angle is implied basic 90: A, B, both, none?

 

[flash3780]

What I was getting at is that measurements have to start from datums (for a cube, A, B, & C), which are perfectly perpendicular to each other. So, in my picture above, the face defined by XX±X would have to be perfectly parallel to the datum that it was measured from at MMC (if the datum was on the LHS, the face being controlled would be on the right).

In CHater's picture, I would say that angle A is an implied basic 90deg angle, and angle B defined by the basic dimensions leading back to the datums.

 

[drawoh]

flash3780,

The datums planes are perpendicular to each other. It would be nice if the datum faces were, but not absolutely necessary.

If the three datums are nominally perpendicular faces, the primary datum is the three points of the first surface, that make contact with the reference face. The secondary datum is the two points of the second face. The tertiary datum is one point on the third face.

JHG

 

[flash3780]

drawoh,
Exactly, the datums are perfectly perpendicular.

When the part is measured it is rested against datum A, then datum B, then datum C. So, in the parallelogram example above, the LHS (if that face is the datum feature) would be brought into contact with the datum and the RHS is defined by dimension XX±X relative to that datum (not from the datum feature).

 

[CheckerHater]

flash3780: "angle B defined by the basic dimensions leading back to the datums"

And what exactly IS angle, defined by basic dimensions, providing that basic dimensions define the perfect shape?

 

[powerhound]

You would have to tolerance the hole pattern with a FCF and that would determine how far off the "point-to-point" angle could be.

Powerhound, GDTP T-0419
Engineering Technician
Inventor 2010
Mastercam X4
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[pmarc]

CheckerHater,

Here is how I see it...

Basic on the meaning of para. 1.4(j) of Y14.5-2009 where it is said that: "A 90 deg basic angle applies where center lines of features in a patern or surfaces shown at right angles on a 2D ortographic drawing are located or defined by basic dimensions and no angle is specified", I would say dimension B from your sketch is meeting the definition.

IMO dimension A can only be considered as implied 90 deg dimension (without word 'basic'). It could be 'basic' if geometrical tolerance was defining the relationship between datum features A and B (e.g. perpendicularity). Since there is no such callout general angular tolerance applies to this angle.

 

[Belanger]

CHater -- you had asked "which angle is implied basic 90: A, B, both, none?"

First, let me take a stab on just angle A:

We can't exactly say if that's a basic 90º because it all depends! If you are talking about the angle between datums A and B, then the answer is always, exactly, basic 90º (per ASME para. 4.1).

But if you are talking about the angle between the two surfaces on the part, then the angle is not basic... it's 90º but must revert to a general angle tolerance.

For angle B:

Yes it is a basic angle per ASME para. 1.4.j.

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

 

[CheckerHater]

Thank you pmarc and Belanger,

I am trying to convince myself that things are the way you described them. Not everything is falling together, but unfortunately I am getting very busy right now.

Will be glad to discuss issue further on next appropriate occasion.