Datum Targets - Best method for evaluating

[[Ev...]]

I have a part with 3 datum targets, A1, A2, and A3. These all exist at different heights, none of them share an X, Y, or Z coordinate. I have no issue measuring these points, but Datum A is used in multiple position callouts. Should these be recalled as feature points into a plane? Recalled regularly into a plane? Recalled into an offset plane? I'm very uncertain on the method to evaluate datum A.

Thanks in advance, happy Friday

[[Al...]]

Hi Evan,

It sounds like you should use an offset plane here. You can add the probing points directly to the offset plane strategy rather than recalling them. When using offset planes you also need to open up the Evaluation window to manually enter the offset for each point.

This thread has some good info: viewtopic.php?t=3754

[[Je...]]

.

There could be some discussion on whether this is an acceptable datum per ASME:

.
.


I suppose you could argue that the "collection thereof" statement qualifies it as a datum, since 3 points is a plane. However, for the sake of your measurement strategy, you should use an RPS (reference point system) alignment in Calypso to constrain your features to datum targets.


.

[[Er...]]

I would try RPS alignment.
A1 A2 A3 would get Z checked.
One point feature XY checked
One point feature Y checked

[[Ev...]]

Please sign in to view this quote.

I'm not very experienced with RPS alignments, I know they're similar to 3D Best Fits. How would I use this? Would I use it as a BA or as a secondary alignment?

I'm currently using A1, A2, and A3 as a 3D best fit base alignment, which hasn't given me any issues. My main concern is that the plane i'm creating between the datum targets isn't correct, which is skewing the reported positions. All of the positions have additional datum callouts besides primary datum A. How does RPS help with this?

Please sign in to view this quote.

Thanks for the info. I'm uncertain of what to use for the offset, can I just arbitrarily pick one of the points to be the "zero" and offset the other two to it?

[[Ch...]]

An offset plane is still derived from the true geometry.

[[Al...]]

Please sign in to view this quote.

If you're using the offset plane as your Z origin in your base alignment, you should set whichever point you want your Z origin to be as zero. Otherwise I think you can choose arbitrarily, as long as your offsets are correct relative to the other two points. Since you're only using three points for the plane you may want to loop the base alignment.

[[Ev...]]

Please sign in to view this quote.

I'm currently using the 3D best fit as my BA, looping 10 times with a break condition of <.002, no issues with this.

Please sign in to view this quote.

So is there any reason to use an offset plane? How should I be using datum targets to create a datum?

[[Je...]]

Please sign in to view this quote.

I wouldn't use an offset plane in this instance. Your best fit alignment is a datum, and it sounds like you constructed it from your datum targets.

[[Ev...]]

Please sign in to view this quote.

So I'm constructing my datum through the best fit alignment, how can I use that as a datum for a position callout? The best fit alignment only creates Datum A and I need to use datum B and C as secondary and tertiary datums.

[[Je...]]

Please sign in to view this quote.

.

Evan, I may have misunderstood. If you have three datums to constrain to, including the three-target datum A, then perhaps you will need to construct A as a plane. I'm having trouble visualizing the characteristics on your print.

My first instinct would be to contact the designer/engineer in charge of this print if at all possible.

[[Ro...]]

Hi Evan,
Without seeing a print it's hard to tell exactly what's going on. Do you have datum targets for the secondary and tertiary datums also? Can you post the print with the datums shown?

Some general info... Datum targets are most commonly used to define a datum from non-machined surfaces, like castings, for example. A standard casting alignment is a 3-2-1 alignment, where 3 points define a plane (3 DOF), 2 points define a line (2 DOF), and a single point (1 DOF). If you have primary datum targets that create a plane that isn't coplanar/normal to your overall part alignment and dimensioning scheme, there should be some sort of basic dimensions defining where the origin is at and how much to offset the points to create a plane that is normal to the part. Sometimes 1 of the points will be 0 and the other 2 dimensioned to it, sometimes all 3 points may not even be the actual datum location. The drawing should tell you this by implied 90 degree angles, basic dimensions, etc. defining the datum locations.

Here's how I would do this...
I would create a stable Base Alignment that doesn't use the datum targets at all. Use the most stable and largest features to locate your part. I wouldn't use a Best Fit as a Base Alignment in this case. I would do the fitting in a separate alignment after the BA. Your Primary datum will either be a simple plane through the datum targets and used just like that, or an offset plane of some sort. Either way, once you create it and your other datums, just populate the Position Characteristic with the Datums you created when you report the results.

Hope that helps.

[[Ev...]]

Please sign in to view this quote.

Not sure why I haven't posted an image of the print itself... See below.

The part is nonstandard with no flat planes anywhere. Every hole has a positional callout to ABC, DBC, or DBE. Datum C is center of the right most threaded hole. Each Datum target that makes up A & D are located on separate Z heights. I had a very difficult time (I'd almost say impossible) creating a standard base alignment and only had luck with an RPS or Best Fit.

Please sign in to view this quote.

Robert, see above for a quick snip of the print. The secondary and tertiary datums are not defined by datum targets. Accidentally cut off Datum C from this screenshot, its the center of the right most threaded hole. Like i've said to Jeff, the part is comprised of nonstandard geometries. I haven't had any success creating a BA off of standard features.



I appreciate any help

1.png

[[Je...]]

.

Ah, medical. That's similar to what I'm measuring now, but I have more planes to work with.

So, are either the top or bottom surface in this view planes?

.
.

[[Ro...]]

Evan,
According to the note on the left, the bottom of the labeled Cbore is the Z origin. So the 3 datum targets, if they aren't already at the same Z depth as the Cbore bottom, will be offset to create a plane at the Z origin.

[[Ev...]]

Please sign in to view this quote.

Neither of them are planes unfortunately.

Please sign in to view this quote.

I've never really had to offset points before, is it as simple as taking the difference between the Z values of the datum targets and the cbore origin, then punching that into the offset box?

[[Er...]]

Please sign in to view this quote.

You could use RPS as either BA or an alignment.

See attached for an example.

Capture.JPG

[[Ev...]]

Please sign in to view this quote.

I see - I guess my question then becomes in this case what's the difference between using RPS vs 3D Best Fit for the base alignment? Realistically if the machine knows where the part is using my method for a 3D best fit, is there any reason to use RPS?

[[Er...]]

[/quote]

I see - I guess my question then becomes in this case what's the difference between using RPS vs 3D Best Fit for the base alignment? Realistically if the machine knows where the part is using my method for a 3D best fit, is there any reason to use RPS?
[/quote]

RPS will only use the axis specified.
As I understand 3DBF will move all points as a group.

As an example. 3 tooling balls used for alignment.
With RPS a total of 6 "axis" must be checked, no more, no less.
Sphere1 set to XYZ, Sphere2 set to Y,Z, Shpere3 set to Z
Sphere1 actuals will always be X0,Y0,Z0
Sphere2 actuals will always be Y0,Z0
Sphere3 actuals will always be Z0
With 3DBF
All of the actuals are a little off from zero.
There may be a way to adjust the best fit that I don't know about?