Cone angle fails in Gage R&R

[Sekhar]

Experts advise required

I am doing a Gage R&R in a part using CMM, which has internal taper profile.The angle is failing in the Gage R&R again and again. All other characters are passing in the same part.
Could any one give some clue to create a perfect auto cone feature. Below is the parameters currently I am using. Any help would be greatly appreciated.

Cone angle 12 degree total tol 0.084
Cone height 3.8 mm
Diameter 60 mm

Parameter

Total points - 3 rows X 9 points= 27
Touch speed 1mm
Azimuth angle 0 to 360 degree
Touch Probe 3X20 ruby ball
Probe module TP20
PH10T


Sekhar

[Schrocknroll]

Short cones like that can be tricky. You may want to try two lines, one on either side of the cone, and measure the angle between, if it's the included angle that you're after. That might give you better repeatability.

[SingularitY]

+1 to the above.
Or you can try concentric circle cone measurement strategies. I've had really good repeatability for cone angles using this.

[Sekhar]

Hi Singularity,

You mean construct a cone from Circles?
or could you please explain bit more?

[SingularitY]

Under the "Auto Cone" feature, there is a tab at the bottom of the box that has different measurement strategies. Here you can select scanning strategies without needing to construct the feature. I'm not sure if these options are available for your CMM probe though.

[UKCMM]

Can you clarify what units the tolerance is degrees, mm, inch, ratio as this will have a bearing on whether the TP20 has the necessary accuracy also is that a total tolerance or +/-.

[KIRBSTER269]

Hey stranger, haven't seen you chime in, in awhile

[UKCMM]

Just trying to keep out of trouble

[KIRBSTER269]

Just like countersinks, cones tend to be based off of a good center. I would try this first to try to get repeatability. Create a CIRCLE at the base of your Cone, as low as you can, without bottoming out. Make alignment to that circle "X" and "Y" Origin. Then create your cone, see if you don't get better readings and repeatability.

[Sekhar]

Nominal 12 degree
Tolerance is +0.042 - 0.042 degree

[UKCMM]

I would say this tolerance is outside of what a TP20 can measure with those limits and IF you could take hits at the max length of the cone. You are only looking at
+/- 0.0015 mm on hit position for that tolerance and given that you cannot use the full length then that figure gets smaller.

You may be able to improve the repeatability by doing as Kirbster269 said and by changing the hit numbers per level to that of the hits taken on the calibration sphere equator.
With that short length you will need to maximise the length of the cone used to take the hits but also be aware that the max length should not be the same or longer than the minimum cone length to be sure of hitting on the cone what depth and ending offset are you using.