-Z Point vector angles aren't square to CMM

[[Ch...]]

Hello all,

I have a program that checks Perpendicularity and upon manual verification, I find it to be much less than what the CMM is telling me. After some investigation, I have been noticing the vectors on my -Z points have been at an angle. 

What could be causing this, and could this be playing a significant role in some of my bad results on Perpendicularity, Runout, etc.? 

Below are images of a sample plane that I had probed on the surface plate of the CMM and the feature data.

 For context I am using a Contura G2 (Vast XT) Calypso 2020. Calibration occurred in April. 

 

Thanks in advance, 

Christian

[[Cl...]]

Had any collisions lately? Is this a new stylus? Is the surface clean and dry?

[[No...]]

The angle of the point vectors do not affect calculations like runout or rectangularity, only the actual point coordinates are relevant for this.

BUT if you get vectors that skewed when touching the granite, there might indeed be something wrong with your probe head. I suppose you took these points in the machine coordinate system?

[[Ch...]]

Please sign in to view this username.

 No recent major Collisions and this is happening with multiple stylus systems. (1mm cone stylus with extension and 8mm with carbon shaft). Surfaces are always as clean as we can get them. 

 

Please sign in to view this username.

 Yes, I was using the machines coordinate system and probing the granite of the CMM. I will reach out to Zeiss Support to try to get this resolved if the head is suspected to be the issue, which is what I was afraid of.

[[To...]]

Can't delete my thoughtless comment, but I can edit it.

Edited July 10

[[Cl...]]

Senior moment? 😉

[[Da...]]

First, The granite can have a porous surface, chips, and scratches. Small probes can fall into the pores. Its always best to check the part, and not a mated surface. When possible of course. 

Secondly, the surface you checked (the granite) is larger than the part resting on it. So depending on how you are asking for the perpendicularity (Is the Perp From the plane or are you measuring the perp OF the plane) it may read as a higher number than when you check it manually. You can make the plane the same size as the part by changing the nominals to match your part size.  I hope that makes sense. 

PSH and WNTGFABS !🍺

[[Ma...]]

I am wondering - are those green arrows from manual or CNC touch? There can be difference if it's show before / after base alignment.

Am I seeing one green arrow with correct vector?

[[Ch...]]

Please sign in to view this quote.

Please sign in to view this username.

 The perp. is from part face to thru-bore.

I probed the plate to verify the vector angles were truly out in relation to the machine and wasn't just the part showing that. This was done in a new file just for testing purposes.

  This was tested this with a 1mm tip as well as 8mm tip. On my parts I am showing upwards of .039" perp. when my manual check (arbor and indicator) is about .0002"(Top end of Tolerance). This is a relatively thin part (.0984"). So, the result is an extremely short Cylinder. With Perp. being that far out, I know it has to be something to do with my program and/or machine, but I am a bit confused as it isn't a consistent result from part to part and all of my related vectors look fine, except the -Z vectors. 

TPISH and WCFOAD! 🍻

 

Please sign in to view this username.

 This was a manually probed test file that I had created in order to show the Forum. The vectors are angled with Manual and CNC runs. Yes, you are correctly seeing that 1 of 4 angles are upright showing a vector that is truly perpendicular to the surface plate of the machine, which only adds to my confusion.