Freeform surface being used in datum reference frames

[[Pe...]]

Hi,
So, I have been seeing more and more prints using a freeform surface as a primary datum in a reference frame that has primary, secondary, tertiary datums. Does anyone have a good method for accurately making an alignment in calypso for this? It has been suggested to me to select the freeform surface and the other two datum surfaces and creating a second freeform from all three, and then creating a bestfit alignment from the result. This method seems a little suspect to me, because I am unsure of where the resultant zero point would end up for this alignment. Would it just be the resultant "center mass" of the points taken, so if one plane had more points taken than others would it be leaning that way? Or, does calypso actually put it at the intersection point of the planes?

[[Ma...]]

I've never seen this type of datum refs. What i saw was 3-2-1 points on drawing defining A,B,C.
But for your problem. I would try as much as drawing allows. Pick only necessary points and freeform surfaces and create a new ff surface. In bestfit restrict translation and rotation as needed to respect refs. and create alignment from it. Then you will create another alignment with usage of rest of points with reference from besfit alignment.

But showing us drawing will be better for better answers.

[[Pe...]]

Sorry, I can't post actual drawing, so I had drew a similar example. Hope this helps

[[Ro...]]

Compound surfaces use the same process as other feature alignments. It's all about how many degrees of freedom the surface is controlling and where it is in the datum precedence. Compound surfaces almost always need a CAD model, unless there is a table of points or something that defines the shape.

For your example (I measured a part that looks exactly like that) you would start by getting some sort of starting alignment that will locate the part in 3d space relative to the CAD model so you can gather accurate surface data. For example, creating circles on the ends for an axis, a width for rotation, one of the holes for origin. The hardest part is usually getting the initial alignment. I would definitely loop the alignment. Then go to town gathering data.Then you need to create the alignments per the print.

In your example it looks like datum -A- controls all 6 degrees of freedom, which means that B and C do nothing unless the print releases DOF and allows them to be controlled by another feature. One way to do this (ASME) is to note the DOF controlled by the FCF and using letters UVW for rotation and XYZ for translation next to the datum callouts. If all 6 degrees of freedom are controlled by datum A then you would do a best fit alignment of the surface data and allow it to fit all DOF. The origin will be the center of the mass of points/surfaces defined by a table or the CAD model. Then you would inspect features to that alignment.

If datum A doesn't constrain all DOF, for example the center line of the part allows it to slide along the axis, then you would create another alignment based on your best fit alignment, and origin on the hole that's datum B and use that alignment to inspect to. I can't see any way that datum C in your example controls anything, but if it did it would be the same principle. Build upon a previous alignment and constrain DOF in order of precedence.

Hope that helps. Let me know if you have more questions.

Robert

[[Ke...]]

Please sign in to view this quote.

Innnnteresting!
Is the freeform surface the actual mating surface?
It looks suspiciously similar to a roof rack component... if someone datumed it wrong 🤣

[[Ro...]]

It looks exactly like a cover for a "military based flying object" that I worked on years ago.

[[Cl...]]

Please sign in to view this quote.

Kind of like a panel on an RAF Lakenheath F-111F