True Position GD&T 1

[Syafiq Hamizan]

Hi Friend,

I need some advice on how to determine x & y coordinate for a true position as per photo.

Thanks.

Regards,
Syafiq

 

[3DDave]

The rules guys will be along shortly to tell you the callout for datum feature A is invalid and must be a perpendicularity tolerance and not a position tolerance unless there is another feature that also uses only datum feature A in the feature control frame for a position or a profile geometric characteristic control.

After that the true position is exactly where it is depicted as being - that is the definition of true position.

If you want to measure how far off the actual manufactured feature is from the true position you need to align to datum feature A, locate the axis of datum feature B, and then clock the coordinate system to pass through datum feature C in the A|B datum reference frame.

After that is set up, translate to where the feature is, [15.00] and set the measurement origin there. With that origin measure where the axis of the feature is; the distance from the measurement origin to the axis is the radial offset; if one is concerned about conformance, look at where the axis is farthest from the true position as measured within the depth of the feature. Double this largest radial amount to get the variation that is used to compare to the tolerance for the feature.

The X and Y position is tougher to determine as the axis may not be orthogonal to the datum reference frame, so you need to decide where along the axis those values are to be measured. It may be at one end the X offset is zero and the Y offset is some value; at the other end of the feature the X offset is some value and the Y offset is zero. This is why it is normal to simply record the maximum diameter of the variation from the true position.

 

[chez311]

The rules guys will be along shortly to tell you the callout for datum feature A is invalid and must be a perpendicularity tolerance and not a position tolerance unless there is another feature that also uses only datum feature A in the feature control frame for a position or a profile geometric characteristic control.

Actually I might be missing something but I don't think we know what A is. The feature with position tolerance to A may very well have a location relationship to A. In regards to the "rules" though a position tolerance with no other location relationship (either other features or the datum feature) may be technically incorrect, it is perfectly interpretable. We would know what the designer meant.

 

[chez311]

Syafiq,

What standard are you working to?

Though position deviation is often reported as (x,y) values, as 3DDave alluded to this does not tell the whole story. Often this is as a result of only sampling at a single Z-level on a feature and totally ignoring orientation variation (aka: assuming zero orientation error). For relatively thin features this simplification may be acceptable but as your relative height/diameter (or height/width) ratio increases it becomes less and less a valid assumption. Position controls both orientation and location and many people seem to forget this.

Just to add what 3DDave already excellently explained, if you are using ASME you may want to look at the new Y14.45 draft ( posted

https://www.eng-tips.com/viewthread.cfm?qid=471922

) for measurement data reporting.

 

[Jacob Cheverie]

Typically, the (x, y) coordinates that are reported would be the Z-level where the positional error is the worst (Y14.45 draft backs that up, I believe). Most of the time that is where the axis is projected to intersect with either end of the feature.

If I have a 3" bore and I measure a cross section that is 1" down, followed by a cross section that is 2" down, an "axis" will be formed from those two cross-sectional center points which will be projected for the remaining inch of the feature. Where that axis meets either end face of the bore will typically be the point that is farthest from nominal.

That's the way that most CMM software will do it. That also does not account for a feature with a large local positional deviation, like a banana bore.

 

[Burunduk]

If I have a 3" bore and I measure a cross section that is 1" down, followed by a cross section that is 2" down, an "axis" will be formed from those two cross-sectional center points

Does this mean that this measurement method is limited to sampling only two cross-sections? The reason I assume so is, that three or more center points will not lie on the same straight line so this may not produce an "axis". Rather it would produce something more similar to a median line (but not precisely so because an unrelated AME needs to be determined as part of deriving a "median line" per definition).

 

[Jacob Cheverie]

Burunduk,

Not at all. That was just an example. You could take as many cross sectional measurements as you'd like. My point is the resulting best fit line will be used as an axis that will determine your positional error when projected, or continued, over the length of the feature.

 

[Burunduk]

I see.
So it's a best-fit line between center points. I suppose no UAME simulation is performed to derive the feature axis in the typical CMM procedure?

 

[drawoh]

Syafiq Hamizan,

ASME Y14.5 has calculations for positional tolerances as an appendix.

Here are my notes on Calculating Locational Tolerances.

--
JHG