Feature evaluation locking vectors

[[Ma...]]

Hello everyone,

i am depeloving a scanning path on external thread to extract axis by external tangential element. Making many measures I found that locking vectors on normal gave me more stable results.

In which occasions do you use that ? Is it a good idea using that on a thread scanning path ?

Thank you.

[[Ma...]]

I am using helix path defined by thread with aid using self-centering option.

Measuring threads were discussed many times with different strategies. Like line scan with many points and filtering and so on.

[[Ky...]]

I would advise caution when constraining the normal vector. It likely gives you more repeatable results, but it doesn't necessarily mean they are correct. I stops Calypso from calculating the axis of the feature you are measuring, setting it equal to whatever it is nominally. E.g. if you are measuring a cylinder with an axis nominally pointing in the normal direction of datum A, then the evaluated feature will be perfectly pointing in that direction.

 

Short cylinders with large diameters often do not calculate their axes very well, which is one reason you would use it, but that I try to avoid using it for that case because usually those should be measured as circles anyway (and circles do this already). The most important reason for it would be for like datum reference frames, making sure that secondary and tertiary datums are evaluated perpendicularly to the primary datum.

 

As Martin mentioned self-centering might help a bit, but you might have some issues if you are using an XXT stylus.

[[Ma...]]

Thank you all' for replies.It was very useful to me.