True Position Question

[[Ni...]]

I have a long plastic tube with multiple diameters throughout the body. I am looking to make sure each diameter is centered in reference to all of them. When i check using true position and the flat side as a second datum, the top diameter is within spec with everything.
When i use true position without any secondary datums, the top diameter is within spec with everything except the last diameter, it is off by a big number.
But
When i use true position without secondary datum from last diameter to top diameter, it is a lot closer.
Just wondering how do i know which number is true and if i should be using a secondary datum or not. Just need to make sure each diameter is right on center with the whole body.
I highlighted the two numbers in question. I would love some help on this.

Thank you

tube.PNG

[[Ma...]]

Datum -A- seems like it could be a troublesome surface to make as a primary. How do your data points look on -A-? Also is datum -B- being measured as a cylinder and if so what kind of strategy are you using? For the Datum -B- cylinder are you using multiple tips to measure this feature? In my opinion it would make more sense to use Datum -B- for the primary instead of -A- but I know the drawing does not state this. Also what kind of machine and sensor are you working with to get your results? A quick check that you may be able to do is throw your features into a coaxial result and see what you get?

[[Da...]]

Looking at you print,
It is a bad GD&T callout you are getting projectional error from Datum -A-
Take the length of the part divided by the diameter of datum -A- that is your projection error in you part.
I would have your engineer team look at function of the part. I don't know with out seeing the total assemble but if Datum -B- is the driving spatial rotation for the part or if Datum -A- is.
Can you give me a total length of your part? I can draw up some examples to show why you are getting different results
Thanks
Dan

[[Ni...]]

Please sign in to view this quote.

Total Length is 9.351.

[[Da...]]

So this is your error
A .005 error on datum -A- will get you a .042 error on the position at the top of the part

Capture.JPG

[[Da...]]

You get a 9X position error from datum -A-
If the part mounts on datum -A- and Datum -B- doesn't control the axis then the print is right
If Datum -B- is the axis then it should be T.P. 0.015 (M) to | B | A |

[[Da...]]

What is the perpendicularity of Datum -B- to Datum -A-?

[[SH...]]

Please sign in to view this quote.

Suppose I use datum B as a cylinder, what would be the calypso algorithm to calculate the DRF of TP. Weather calypso use origin of the local feature coordinate of datum B or project point to the plane datum A ??

[[Ni...]]

Please sign in to view this quote.

We are using a Contura with a Vast XT head. 1.5mm ruby probe.
A seems good? I just scan the surface as a plane.
B is being measured as cylinder at multiple spots on the tube. Only using 1.5mm ruby for everything.

How do i throw my features into a coaxial result?

B.PNGA.PNG

[[Ni...]]

Please sign in to view this quote.

By the part mounting on Datum A or B. Are you referring when its being cut?
or loading on cmm to measure?

Perpendicularity for Datum B to Datum A is called out on the print at .005
And i get .00257 in calypso.

[[Ni...]]

Also just a thought.
.325 to 1.116 AND datum A is within spec.
Without Datum A, it goes out of spec. by .200

1.116 to .325 AND datum A is within Spec.
Without Datum A, it goes out of spec by .001

If i only take into consideration the results that have Datum A included, the part shows to be pretty straight when comparing true position from the diameters at the ends.
Would that be accurate?

What about the results that don't include Datum A. Are those results wrong and something we shouldn't worry about?

[[Ma...]]

Are you only measuring partial arcs or are you doing full 360 degree scans for your Datum -B-?

[[Ni...]]

Please sign in to view this quote.

Only doing partial 180° scans on Datum B.
I did recheck the parts multiple times by taking them out and loading them again on the CMM and the diameter for all features were off by around ±.0001 each time. Same results on True positions too.

[[Da...]]

Are you checking the last feature as a cylinder or circles?
Perp looks good for datum -A- to -B-

[[Da...]]

Please sign in to view this quote.

Per the Print the results would be wrong. But there is something more going on. When you check Datum -A- to Datum -B- are you filling in the fields like this?

Capture.JPG

[[Ni...]]

Please sign in to view this quote.

I am checking the last bottom feature as a cylinder. I have also constructed circles from that cylinder feature and used that in the true position. Changes the results by a few tenths.

TP Bottom.PNGperp.PNG

[[Da...]]

You are getting projection error from the axis of the feature.
If your end feature is off by .001 to that axis of Datum -A- setting at 179.81°
Then you get a position of .030 at Datum -A-

What you are going to want to do is create Points to check it
Try first to check the position with circles at 3 or 4 different heights on the last feature.

Thanks
Dan

Capture1.JPGCapture.JPG

[[Da...]]

What is the result if you check coaxiality from feature to datum -B-?

[[Ni...]]

Please sign in to view this quote.

I'm sorry i do not understand how to check using points.
Coaxiality is attached. Wasn't sure if i should use MMC on top diameter feature but i did.

Thank You

coax.PNG

[[Da...]]

Create Theoretical Plane Feature.
Using the tool Intersection
Create the Intersection of Feature to the Plane.

Then check the True Position of the intersections

Capture.JPG

[[Da...]]

Your true position of the Feature is almost the same as the coaxiality its looking like your part is bad with out MMC
What is the number with MMC on the True Position?
That will be in your report