Base Alignment for Tube

[[Cl...]]

Don't know how much I'd trust a 180 degree circle?

[[Da...]]

Please sign in to view this quote.

I would 100% agree with that also

[[Cl...]]

You could try coming in from Y+ & Y- using linear scans in a
cylinder. At least that way you'd get 360 degrees of coverage.

[[Ri...]]

It's not going to matter what the Base Alignment is using when you tell it to use Datum A as the Primary, and Datum B as the Secondary. Those two need to be flipped around to provide a stable measurement. I would speak with the Engineer and talk to them about this.

[[Ni...]]

Please sign in to view this quote.

So create a line throughout the part on Z+, Y+ and Z- of the part, using my Y- probe. Then using Y+ probe I make a line on the Y- of the part?
Would that be enough or better than creating the 180° degrees on 3 spots?

[[Ni...]]

Please sign in to view this quote.

Could you please provide an explanation on why they need to be flipped? All programs have been following the print set by the customer so far so it will be new to flip the datums on just this part number.

[[Ri...]]

Datum A is going to setup your Spatial Rotation and your X axis origin. Datum B is going to setup your Y, and Z origin.

Now, you have to imagine how small the surface area of Datum A is, and then also imagine how far away the Ø.325 is away from Datum A (roughly 9").

My suspicion is that their intent is to assume that the Ø.325 is coaxial to Datum B (which is much longer, and much more stable of a Spatial Rotation).

Either way, play around with the True Position, and see how much of a difference in results you see by flipping the Datums around.

[[Jo...]]

Please sign in to view this quote.

What Richard is talking about.. Imagine that you were manually measured the part by placing Datum B in rollers and spun it with indicator on Datum A face to measure axial runout and got .005" . Now stand that part on Datum A and spin on a turntable(like a record player orientation). It would have a tremendous runout down the length of Datum B. It is the same reason you base alignment off largest area.

[[Ni...]]

Please sign in to view this quote.

So how exactly would my base alignment be?
I was told to create circles on the end of the parts and create a 3d line and use for Y and Z origin.
Then use Datum A as the Spatial and X origin.

How would I go about using Datum B in the base alignment? or how would I create a 3d line for datum B if is measured as a cylinder.

[[Jo...]]

Please sign in to view this quote.

Your alignment is fine and when using Datum structure shouldn't influence the position. Do as Richard suggested. Report Perpendicularity of A to B, and then B to A and it will become apparent, they are NOT the same.

[[Ni...]]

Please sign in to view this quote.

I do get a difference when I flip them. Also regarding the base alignment. Would Datum B for my Y and Z origin, would that be measured as a 3d Line? Cylinder? I have been given different ways to do it so I am not sure what that should be.

[[Ri...]]

Please sign in to view this quote.

All of the suggestions for setting up your Base Alignment are valid.

My comment refers to when you setup the True Position characteristic. When you build the DRF, it isn't affected by the Base Alignment it creates it's own "alignment".

[[Ni...]]

Please sign in to view this quote.

What do we do if they are not the same?
Is that when we go for switching Datums and start measuring the true position of the .325 diameter to Datum B as primary and Datum A as secondary?

[[Ni...]]

Please sign in to view this quote.

Got it. Then we should switch the datums on the primary and secondary?
Also.
I checked a part and the true position of it was .020. I rotated the part 180 degree and measured it again and it was .015.
is that because we used Datum A as the primary?

[[Jo...]]

Please sign in to view this quote.

That's what Richard is getting at, but design intent needs to discussed with engineer, or at least disclose your method. If there is not someone readily available you could report both with disclosure, one is per print, and the other is swapped. Projection error is going to kill your position results with Datum A as primary.

[[Cl...]]

Yes, but not a line, a cylinder. Pick the cylinder from lines icon in the
cylinder strategy (top row, sixth from the left). I would use at least 6 lines.

[[Ni...]]

Please sign in to view this quote.

I have results for both per print and Datum B as primary.
Per print(Datum A Primary and Datum B as secondary) I get around .025'' true position
Flipped(Datum B primary and Datum A as secondary) I get around .015'' True Position.

Would the flipped Datums be the correct way to measure this? It is something engineers are asking and are not sure themselves.