Straightness

[LoriBru]

This is my first post although I have read many treads. I have a fcf with a callout for straightness in a cylindrical tolerance zone @ 0.1mm @ mmc and it's datum A. Datum B is a notch at one end of the part. It's a pivit pin that's diameter is 19.80 ±.05 and length is 283.90 ±.13. First I put it on the surface plate and measured it with an indicator. The QE wants it measured on the CMM. I have a manual DEA Swift with PCDMIS 3.7. I measured 10 circles from one end to the other, constructed a line through the two at the ends, leveled at the notch, zero'd through the line and asked for X,Z for all of the circles. Is this the best way to measure this? Any feedback will be much appreciated.

Thanks
LoriBru

[Schrocknroll]

Construct a 3D line through all the circles and just dimension the straightness of the line. There is a straightness callout just like true position and profile. That should get your answer.

[John Kugler]

Did you try creating a cylinder from all your circles? Then dimension straightness of the cylinder.

[LoriBru]

I did contruct a 3-D line. The straightness measurement gave me 0.000. I did not try the cylinder, I will try that. The X and Z for all of the circles are measuring +.160/-.257 Z and +.057/-.057 X. The surface plate and indicator gave me about a .15 TIR. I'm just not confident in a way to measure this.

Thanks

[Matthew D. Hoedeman]

Is the line constructed using ALL the circles you measured? If it is only using the end circles (just 2) it will always give a perfect result.

[LoriBru]

It was only through the two end circles. I constructed through all ten and got a reading of .366. I'm going with that.

Thanks

[kbotta]

There are 2 types of straightness. Straightness "line element", and straightness "axis" (also know as the renegade...).
Being that your fcf is associated with a FOS (feature of size) you have the latter.
One interesting note I'd like to make is, barring any note on the drawing, this fcf and feature is the exception to the taylor principle (or rule#1, or envelope rule...).

Sounds like you are on the right track. Did you try measuring as a cylinder?

KB

[LoriBru]

I constructed a cylinder from all ten circles. I get 0.000.

Thanks for all your help.
Warning, I'll be back

LoriBru

[kbotta]

Try measuring as a cylinder instead (not constructing one from circles).
Also, were you in the correct working plane?
kb