Questions on High point plane V2016

[BKulpa]

Back ground is, we have a 2 tooth row sprocket this is 23.5mm tall with a D shaped center bore. The flat of the D shape is dimensioned as angularity 0.05mm -A- -B- -C-.
-A- is one of the flat ends
-B- is the cylindrical portion of the center bore
-C- is the center line of the timing tooth on one row
The part is sitting on a 6" square magnet for -A- and a plane is measured and leveled
6 linear open scans in bore that are then filtered for construction of a cylinder for -B-
2 center scans in the gap on each side of the tooth to construct a mid point for rotation of part to -B- -C-
Alignment is level Z axis to -A-
Rotate to Y axis -B- & -C-
Translate Z axis to -A-
Translate X&Y to -B-
The flat in the bore is measured with a patch scan of 6 rows and a filtered plane constructed.
Our problem is there is too much form error in the flat causing the angularity to be OOT, which is correct to ASME Y14.5 using exact measure.

I am trying to use a high point plane to establish a functional dimension without including the form error just the 3 points that the shaft will engage on.
Is there a way to easily extract the 3 points that the HP plane function uses?
Why is there flatness error in the constructed HP plane?
Why is my angularity dimension of the HP plane zero?

Our supplier is using V2019 and exact measure and is not seeing the form error as much as we do.
I've verified the form error on the surface plate with an indicator by leveling the flat surface and traming down the length of flat and seeing >0.04mm TIR

[JEFMAN]

Hi BKulpa ,
Many questions in this thread... !

Extracting 3 points should be done with a assignment. Calculate the distance between the high point plane and each point of this.
ASSIGN/V1=DOT(PL1.HIT[1..PL1.NUMHITS].XYZ-PL_HP.XYZ,PL_HP.IJK) (array of distances - the three lowest should be zero)
Then ASSIGN/V2=MININDEX(V1)
and construct 3 generic points with PL1.HIT[V2[1]].XYZ, PL1.HIT[V2[2]].XYZ,PL1.HIT[V2[3]].XYZ.

I think the flatness is not zero because you use all the hits, but maybe the angularity is zero because it's a constructed plane (not sure of this...)

Your supplier can find different results for many reasons (matybe he doesn't use the same fixturing, so the magnet doesn't constrain the part, the V2019 tangent plane is really different than before 2018R2, the probe head isn't the same, tip diameter different, cmm different...)

Hope this helps...

[bfire85]

Off subject, but would angularity be dimensioned to three datums? Angularity is just perpendicularity at an angle rather than 90°. Maybe there is some calculation error in dimensioning this against 3 datums in Xact. What happens when you measure it against just A?

[BKulpa]

The angular relationship to the sprocket teeth needs to be maintained to the mounting shaft.
And when I dimension it to -B- & -C- the results actually get worse. Same with legacy dimensioning.

[BKulpa]

So I had a part measured this morning (only one tech allowed in the lab at a time) exact measure LSQ plane;
Angularity to -A- -B- -C- result =0.0495mm
Angularity to -A- result = 0.0462mm
Angularity to -B- -C- result = 0.0587mm


Angularity High point plane result = 0.0000mm

[JEFMAN]

I think it's a construction problem...
In ISO world (and I believe ASME the same), B should be perp to A so the angularity shouldn't change.
Constructing B perp to A needs projecting kits on A, then construct the max inscribed circle and define B on it.

[Liambo734]

I know in v2011 the high point plane didn’t work but they reprogrammed it in v2013, I didn’t test it in that version.

I always had an issue with constructing on the 3 highest points, the issue is the 3 highest points can be very local to each other and therefore lead to a very weak primary datum.

I found that if I detected the highest point on the measured plane and projected the best fit plane to that point that it was more stable for me.