Super easy Calypso 101 Question, 3d best fit Datum A

[[Ro...]]

We have a particular customer that send us jobs with very few flat surfaces. Everything is curved profile surfaces, including the top A datum.
After a conversation with the customer on how they inspect the parts, they said they take about 50 points on the curved A surface (depending on size) and use a 3D Best Fit alignment.
Datums B and C are generally drilled holes.
How do you set a profile to 3D best fit?
Calypso help is not very helpful, and searching the forum for 3d best fit is giving me a lot of bore pattern TP's, not profiles.

[[Ma...]]

How it's stopping rotation? Using 3D BF just from one curved shape is not good.

I would use something to get correct position, then using 3D BF on that shape with only move translation. But it depends on drawing how it fits together.

[[Ri...]]

Probably easier to just use Free-Form, turn on the Best Fit, and then inside of the Free-Form you can tell it to Create Alignment from the Best-Fit.

And then if you wanted to make your Datum B and C alignment, you could leverage the Geometry Best Fit, and have it live inside of the Free-Form Best Fit Alignment. You'd create one for Datum B to control those specific degrees of freedom, and then you'd create one for Datum C that lived in the Geometry Best Fit alignment from Datum B.

A lot of steps, and would require testing, but without seeing your part/application, it's hard to know exactly would be the best course of action.

I've had a few applications that had similar-ish alignment requirements, but they are all unique.

[[Ro...]]

Please sign in to view this quote.

So does this make sense to you........
-Open new FF surface and recall feature points of several points on the datum surface.
-In the Eval section, turn on Gauss (noted by customer) and create an alignment.
-Create a secondary alignment and recall the newly created alignment as the base, then enter my physical B datum into the X and Y, and a created 3d line between B and C into the rotation.
Does this alignment make any sense to you? Because its pretty similar in results with my previous attempt.

dfejkgfhdsfkljghwklj 1.PNGgfghdfskljghslkfdgh 3.PNG

[[Je...]]

Please sign in to view this quote.

.
Roberto, are you trolling us with the "Super easy Calypso 101" title 🤣?

This is advanced alignment stuff. Thanks for sharing about the situation. Please keep us posted on how it turns out.
.

[[Ri...]]

Please sign in to view this quote.

You can try that method, but be careful. I have had to learn the hard way of using a secondary alignment that references another alignment. In that case, what we learned was that if you change the spatial rotation, it wipes the referenced alignment entirely, and starts pulling from the Base Alignment.

In your case, I think you'd be OK, but I was talking about using the Geometry Best Fit, because in that one you can absolutely control what degrees on freedom are being changed in respect to the referenced alignment.

But either way, you're on the right track - at least on my initial thought on a way to tackle this.

[[Ma...]]

Richard got a point - if you want to be safe, then make theoretical 3D line in your main axis with feature alignment from FF. This will prevent unwanted glitches.

[[Ro...]]

FYI, in case anybody was wondering or following, the method above works well for creating an alignment when Dat A is a non standard surface.
I actually got to use it again this week when my Dat A was a small section of a giant cone, and B & C were spheres.
I input the standard ABC into the Profile DRF, and the arrows were wildly out of tol.
I rescanned the bottom cone with several Free Form scans, then did as shown above, then created a Secondary ABC alignment. When i put that into my profile everything went nice and i was in tol as expected.

[[Ro...]]

Please sign in to view this quote.

It works really well, see my previous post.
This will definitely be added to my bag-o-tricks.

[[Ma...]]

Please sign in to view this quote.

quoiting to find later