run out

[pedrof]

Simple one for you gurus out there. Is the perpendicularity of datum G to datum H implied by the total run out ( the 0.02 would not be achieveable if the surfaces were not perpendicular)? OR does the H datum need to be controlled to datum G with its own geometric tolerance?


I have attached a jpg.

 

[pedrof]

not sure if it attached.....

 

[TheTick]

Using two datums on a callout does not imply any control of one by the other. So... "B"... H datum needs to be controlled to datum G with its own geometric tolerance.

 

[powerhound]

Runout to datum H as secondary has no effect here. It should only be to datum G. You can call runout to H as primary and G as secondary.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[TheTick]

Powerhound: I had to dig a little for this. I found the same picture in another text here

http://www.ohio.edu/people/tc285202/288/runout.pdf

The callout is valid. The "H" datum only acts as a stop, much like a tertiary datum in the archtypical 3-datum position callout. Otherwise, "H" doesn't really influence the runout.

 

[pedrof]

Having said that though, a similar example Fig 9.5 in ASME shows all surfaces referenced to G and H (or D and C in the example) and says "at any measuring position each circular element and each surface must be within specified runout tolerance when part is mounted on datum surface C and rotated 360 about axis D"
So looks like the datum order is not that critical but worth including the secondary since it would be used in inspection.

 

[powerhound]

Tick,

Imagine if the datum H was NOT referenced in the FCF. What would be different about how the runout is checked? Would it yield a different result whether datum feature H was against a stop or not?

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[TheTick]

About the only difference would be potential variations due to part clamping differently.

Personally, I wouldn't use the second datum. But, it is still valid.

 

[powerhound]

I guess it depends on how we choose to define "valid."

Since datum H has absolutely no effect as a secondary datum, then it should not be there.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[Belanger]

Datum feature H doesn't need a perpendicularity control of its own to datum G. Even if H and G were not perpendicular, the datums referenced instruct us to establish axis G and then push it to contact the highest point of datum feature H (in a perpendicular direction).

I don't think it's wrong to leave H on there. There could be a situation where I need to leave it because I want to imply a simultaneous requirement with another feature control frame that would reference those same datums in that same sequence. Just a thought.

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

 

[powerhound]

I don't think a simultaneous requirement would ever apply though. It only applies to features that are located with basic dimensions. That means position and profile. While the axes may be basically oriented, they are not basically located.
I still don't get how datum H as secondary makes one iota of difference.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[Belanger]

Yeah, I guess I agree -- the simultaneous thing has no place in a runout check. Shrug.

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

 

[axym]

Hi All,

Interesting discussion. This is one situation where we have to be fastidious with terminology.

I think we all agree that referencing datum features G and H in the same feature control frame does not control the perpendicularity between these features. The datum feature simulators (contracting chuck for G and flat plate for H) and the datums (axis for G and plane for H) are defined to be exactly perpendicular to each other, but the datum features are not. The perpendicularity between feature G and feature H would have to be controlled by separate tolerances on those features. The datum reference frame |G|H| could still be established even if the features were significantly out of perpendicular.

I re-read the original post, and there is something else that has not really been answered. The geometric tolerance itself has no connection with on the allowable geometric error in the datum features. In other words, the Total Runout tolerance of 0.02 is still achievable even if the datum features have perpendicularity error (or flatness error, cylindricity, whatever) that is greater than 0.02. It is a good idea to apply tolerances to the datum features to avoid rocking and instability, but it is not a requirement.

I agree with powerhound that datum feature H has no effect when referenced as the secondary datum feature in this FCF, and thus should be omitted.

The issue of simultaneous requirements with runout tolerances is an interesting one. Y14.5-2009 excludes it - Section 4.19 specifically states that sim reqts "applies to position and profile tolerances that are located by basic dimensions". It is possible to envision applications where simultaneous requirements would apply to runout tolerances or even to orientation tolerances, but the standard hasn't gone there yet.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[CoryPad]

From a measurement standpoint, doesn't adding H to the FCF imply that diameter G is fully engaged in a chuck or similar fixture? Without H, couldn't I grab just the first bit of diameter G and measure? If so, any straightness error along diameter G may or may not be measured depending on how the part is chucked.

 

[axym]

CoryPad,

G needs to be fully engaged along its entire length, even if H is not referenced.

Evan Janeshewski

Axymetrix Quality Engineering Inc.

http://www.axymetrix.ca

 

[pedrof]

I think it's clear that datum H as a stop for the part in the chuck doesn't affect the diameter runouts.
We have a culture here of labelling not just the diameters but also the faces parallel to H with the same sequence in the frame. i.e. all frames basically read runout to G and H.
My second image I attached was trying to indicate that for the flat faces the order should be changed, in fact the second datum could be left off altogether as in the other case.

Axym - what about the statement I found from Fig 9.5 in the standard, that seems to link the orientation of the 2 datums?

 

[powerhound]

pedro,

Evan hit the nail on the head when he said that we need to be correct in our terminology. Datums will always be oriented exactly with respect to each other, datum features will not. The diameter identified as D is the datum feature. This means that it is this feature that will be used to establish datum D, but the diameter itself is not datum D. The same thing goes for the plane called datum C. The surface will be used to establish datum C but the surface itself is not datum C.

John Acosta, GDTP S-0731
Engineering Technician
Inventor 2013
Mastercam X6
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[3DDave]

Hi Tick -

Love the PDF. I wonder where in the standard is the interpretation of "pressed tight" as shown on page 7. Is there a difference in interpretation if it is only pressed pretty hard, or very hard, or just barely touches?

It's the sort of diagram someone in manufacturing might create because that's how they set it up, not realizing that it's a meaningless addition to establishing the DRF. Maybe they had someone chuck a part on just the first 1/32 in and they feel this is a way to force it to be fully engaged, in spite of the definition of the entire surface as the datum feature that already requires it to be as fully engaged as practical.

Still wrong.

Unless there's an internal standard on mentioned on the drawing that says what to do with callouts that exceed what's in the standard; then go for it. Whatever a company wants to do is fine as long as they document their own brand of crazy.