Profile or True Position back to itself

[[Jo...]]

Something I frequently encounter on customer drawings are true positions and profiles with no datum reference. The snippets below show an example of one part. In this case, for the .001 profile I would create a geometry best fit alignment and select all of the applicable basic dimensions and make that the datum reference frame for each profile characteristic. In the second snippet, each individual corner of the pocket has a .0015 profile with no datum reference. The proper solution would be to create its own geometry best fit alignment selecting just the feature, and using that as the DRF for the profile? What I've done in the past is use the feature itself as the primary, secondary and tertiary datums (I'll probably get roasted here for this) for the profile if it is a single feature, but having the feature and all three datums be the same seems kind of dumb. I imagine it would be like measuring the length of a yardstick using itself to measure it with.

As a rule of thumb I now use Outer Tangential Element as the evaluation method on all datum features and LSQ on the measured features, as I understand this best simulates form fit and function. This has yielded more favorable results during inspection when utilizing GBF alignments but I do worry about the "too good to be true" factor. When tolerances are generous this isn't an issue, but when working four and five decimal places out it can make the difference between green and red when the part is done running.

Any input or suggestions on my question above or the methods I've described are greatly appreciated.

[[Ri...]]

If you have Free-Form, I would try using that instead. Then inside of the Profile, you can select "No Datum Reference".

[[Jo...]]

Please sign in to view this quote.

Unfortunately we do not have free-form on any of our machines.

[[Za...]]

According to your print looks like the profiles should be unconstrained (no datum reference). So the easiest common solution is to put the plane feature itself 3 times for the datums. Refer to the screen shot.

If it is a freeform feature then you can actually select the drop down and do “No Datum Reference”.

[[Ri...]]

Please sign in to view this quote.

Understood. I believe what you are doing is your best bet if it is a multi-feature evaluation.

[[Me...]]

Why not just include the corner radii in the geometry best-fit with the other features? If you’re worried about the points in the corner influencing the fit, leave them out, but evaluate the .0015 profile control of each corner to the GBF alignment.

[[Jo...]]

Please sign in to view this quote.

I'm not worried about how the part functions or fits, that's the customers issue to worry about. I am concerned about the parts meeting print and since the rads are specifically called out separately, with their own individual profile callouts, I wouldn't use other features as datums for them.

[[Me...]]

Please sign in to view this quote.

I was referring to the fitting of the associated feature(s) & best-fit alignment. 😃

Just, be aware that short cylinders established from partial arc segments can come back to haunt you. This is especially true if you’re using them as datum references. Hopefully you and your customer can work out an appropriate method.

[[Er...]]

Correct me if I am wrong, but Profile without a DRF is just the form of the surface. Take enough points on the surface and you should be able bring in the form of the surface and call this measurement the profile.

[[Jo...]]

Please sign in to view this quote.

Not knowing anything about Free Form, I would venture to say that is probably where that add-on would come in handy because it would give you the option to have no DRF. Essentially, no reference to anything but itself.

[[Ri...]]

Please sign in to view this quote.

And size.