Base Alignment vs DRF

[[To...]]

I have a disc shaped part with Datum feature A is a plane on the top, Datum feature B is a the ID Cylinder and Datum feature C is a cylinder is rotated 119° BASIC about B from 3 o'clock position. However, the OD (not shown) is a cam shape with Line and Surface Profiles to BC and BCA.

Originally, I used the
Spatial: Plane_Datum A
Planar: Int_C (intersection of Cylinder_Datum C and Plane_Datum A)
X Origin: Cylinder_Datum B
Y Origin: Cylinder_Datum B
Z Origin: Plane_Datum A


After thinking about this for a bit, I think I would like to change the Spatial feature to Cylinder_Datum B and use Cylinder_Datum C for my Planar. How would I go about creating the related actual mating envelope of Cylinder_Datum C relative to Datum B? Even if it means I still have to use the intersection of Cylinder_Datum C and Plane_Datum A.

Am I over thinking this? I really want to measure the OD cam normal to B and C due to very tight tolerances.

BA vs DRF.JPG

[[Da...]]

You are aware, that you should never use the base alignment as the datum structure in characteristics but rather use individual datums? The base alignment NEVER uses mating envelopes on the used features. If you insisted on doing it that way, you could create a feature of datum C with restricted degrees of freedom to B as an outer tangential element, but that’s one more step.

Better make the profile evaluation of your cam shape not to the base alignment, but directly to B as primary, C as secondary, and B as X/Y and A as Z tertiary. The related actual mating envelope will be automatically used when you have your ISO 5459 (if you have an ASME drawing, don’t mind the name) setting activated.

Actually, come to think of it, one significant difference between ISO and ASME that applies here, is that you need to restrict the translational DOFs for datum C in X/Y for the RAME to keep the distance to B fixed (at C’s basic distance to B, and for that, all nominals have to be correct), which does not apply to ISO.

[[To...]]

Please sign in to view this quote.

Yes, I am aware.

My goal was to establish the base alignment so the measurements were normal to the important feature surfaces. In other words, if I use the top plane for spatial and it was not perpendicular to the cylinder, when I measure the cylinder, the scan would not be normal to cylinder (unless I added extra steps). So, by measuring the cylinder as the spatial, the scans on the cylinder and OD cam would be more normal, hopefully giving me the best results when I used the cylinder for the Primary datum in the profiles.

My question was really directed to the planar rotation. Since an intersection of cylinder C at plane A would not represent the RAME, I was looking for suggestions on how I could best establish the Base Alignment to mimic the DRF BCA as closely as possible, specifically related to datum C cylinder as the planar.

Like I said earlier, maybe I was overthinking it. However, we were trying to make the most accurate part possible and then measure it with the most accurate methodologies.

See my other related post on the results of the test.
viewtopic.php?f=13&t=5327

[[Da...]]

Okay, to get the RAME to use in the base alignment, work with constrained degrees of freedom when recalling the feature points into another feature and then use that feature, or you could use the recalled feature points to construct a circle in the same direction as the datum cylinder and as an outer tangential element. That circle then has the same coordinates as the RAME.

[[Ma...]]

I am missing some terms, but i would use datum C as ASME uses ( outer / inner tangential - select which is better for hole ), and use midpoint for rotation.
This will give you perhaps the most stable results.

I just gess because for best results is knownledge of usage of a part crucial.

[[Er...]]

With the DRF calling for BCA I would use an alignment that was applied in that order. That way the vectors for probing would be as correct as possible.

For me on Calypso 5, any construction in an alignment gives random issues. For -C- I would use a circle for planar.

If the program is already written, I would make alignment BCA. Then measure the features again in that alignment for the characteristics. The Base Alignment features would just be for finding the part.