# Measuring complex Position features / position to 3 datums

• 03-31-2021, 12:51 PM
inkology
Measuring complex Position features / position to 3 datums
I am a new QT w/5 years working in inspection, though I am doing QE work. I have no formal training in GD&T, inspection, or engineering. Just OJT & lots of self study. I took some machining classes once upon a time but have never been a machinist. I have basic programming skills in PCDMIS & better than average programming skills on a vision system. My new duties have removed me from furthering my programming skills on either.

I am in an argument w/manufacturing engineers on [B]measuring [/B]position to 3 datums in multiple working planes. I hope my words can paint a decent picture...

I have an object where the position of a bore needs to be calculated using datums that are on different working planes. call out is - position |Ø.010mmc|A|B|C

Datum A is a plane that is co-planar to the axis of the bore (I use this measurement for X)
Datum B is the center of a bore that is through Datum A... After zeroing on the center of Datum B I need to rotate the part 90° to measure to the center of the bore if I am using an electronic height gauge (This is Y)

Datum C is normal to the axis of the bore & would be the plane that the bore is milled into.

My understanding is that position is 2D... the position formula is for x & y. the tertiary datum is used to restrict freedom for fixturing purposes & if it it applies to anything from a measurement standpoint it is asking that the feature be perpendicular to the ref datum (C).

Based on that thinking -

I say I can measure/calculate position in X & Y & the resulting number is a valid position number. I say I can measure this on an electronic height gauge. I can account for the 3d nature of a bore (cylinder) by measuring both ends of the bore via EHG from the same datum ref & if both are w/in the tolerance then it is passing.

I am being challenged that this is wrong because of datum C. therefore it can not be measured via something like an EGH. but only a CMM.

a secondary question in this particular matter is that the CMM program only constructs a circle from the bore - not a cylinder. therefore perpendicularity to C cannot be established & the CMM is doing the same thing as the EHG taking measurements from one end of the bore. so if i am taking 2 measurements on the EHG - it is actually more accurate than the current CMM program in regards to the feature being 3d.

a third consideration is that Datum A is often overbuffed at the machine... changing the vector. the CMM is drawing a line & rotating to this line. this line is projected & the measurement is drawn perpendicular from the bore to the projected intersection of that line & the bore. this can be a vastly different measurement than taking a single point from datum A (as you would on an EHG). & this is actually where the argument is arising. I argue I can take a part that fails this position measurement on the CMM & verify it as good via EHG. Primarily because Datum A is changed & because the EHG allows for 1 point to ref from. (i have suggested that a point on datum a be used to calculate the position with on the cmm as well).

Could someone shed some light on (via answer or link) measuring more complex position call outs please (i.e. more than a hole in a square plate).

Specifics in helping me argue my position -if I'm correct, or explaining why I am not would be welcome.

thank you.
• 03-31-2021, 01:29 PM
Kelly Bramble
[quote]My understanding is that position is 2D[/quote]

No...

I think this is the second question in a week regarding the interpretation and setup for measurement to datums. Position tolerance can be 2D in very thin cross sections like a .0005" shim however if there is a 3D Datum Reference Frame then the interpretation of the geometric tolerance boundary for position (and others) is 3D.

My best and most practical answer is that you need significant training in the interpretation and application of datums and the perspective geometric tolerances.

I'm not going to attempt the train you in this forum as you need a lot of basics..

If you have a premium membership you can access my last book on ASME GD&T per. ASME Y14.5-2009 "Geometric Boundaries III, Interpretation and Application of GD&T per. ASME Y14.5-2009 and ASME Y14.41-2012".

[URL]https://www.engineersedge.com/Engineering_White_Papers/Dimensioning_and_Tolerancing/geometric_boundaries_iii_14849.htm[/URL]

You can also purchase my latest and greatest Geometric Boundaries IV on this website and have it delivered to your location.
• 03-31-2021, 02:28 PM
inkology

[QUOTE=Kelly Bramble;17280]No...[/QUOTE]

Then why does the position formula only account for x & y... not a z?

--------------------------------------

attached is a rudimentary sketch of what i am trying to measure. again... my argument is that i can capture x & y with an EHG & use position formula.

again - the cmm program is not constructing a cylinder from the bore - but rather a circle.
• 03-31-2021, 03:18 PM
Kelly Bramble
All GD&T Tolerance boundaries are applied to the full depth, width and length of the feature.

X and Y measurements you're referring to are simple 2-d measurements from a specified and established Datum reference at any cross section along the Z length. 2*sqrt(X^2 + y^2) = tol is converting a 2d old fashioned measurement method into a diametric true position at THAT CROSS SECTION. This measurement and calculation is done at many-many cross sections to verify the 3D tolerance boundary start to end of the feature. Most correctly, you should be measuring from the mating envelope if this is a tight toleranced part.

Functionally, you may actually need a qualified check gage.

Rule #1 also called the Envelope Principle <- learn it

If you prefer a 3D equation:

Tolerance = 2 * sqrt(x^2 + y^2 + z^2)

[SIZE=3][B]This is basics - again, not training you here.[/B][/SIZE]

[COLOR=#000000][FONT=monospace]Geometric Metrology, Dimensional Tolerances Inspection and Practices in Manufacturing Fundamentals[/FONT][/COLOR]

[URL]https://www.engineersedge.com/geometric-metrology.htm[/URL]
• 03-31-2021, 05:34 PM
Cragyon
Nothing worse than a poster that tries to change the question and claim the responder is "[COLOR=#1D2228][FONT=&amp]live in a theoretical perfect world".

"[/FONT][/COLOR][COLOR=#333333][FONT=Verdana]is that the CMM program only constructs a circle from the bore - not a cylinder."
[/FONT][/COLOR][COLOR=#1D2228][FONT=&amp]The CMM is not being utilize (programmed) correctly if only a single circle is being measured. Measurements should be taken along the entire length of the hole feature not at one location. 5x rule - ever heard of it?

"[/FONT][/COLOR][COLOR=#333333][FONT=Verdana]a third consideration is that Datum A is often overbuffed at the machine... changing the vector"
[/FONT][/COLOR][COLOR=#1D2228][FONT=&amp]Datums are established from the tangent mating plane by measuring from a datum simulator that the datum is placed against. For lower toleranced parts collecting a grid of measurement locations effectively nulling variations in the datum surface.

"[/FONT][/COLOR][COLOR=#333333][FONT=Verdana]I argue I can take a part that fails this position measurement on the CMM & verify it as good via EHG."
[/FONT][/COLOR][COLOR=#1D2228][FONT=&amp]Kelly calls this "Smoke and Mirrors measurement". Your probably passing a part/feature that should fail - what you're doing is cheating a part to pass.

[/FONT][/COLOR][COLOR=#1D2228][FONT=&amp]A height gage should measure to the actual mating envelope established by a gauge pin in your example - Kelly showed me that 25 years ago back at the real world company we both worked at.

There's a lot of wrong going on in your measuring method and argument. [/FONT][/COLOR]