Datum Targets

[[To...]]

I'm curious what type of points and strategies you guys use when you have a Datum Target with a diametral zone or rectangular zone. What if these targets were on a cast surface with a draft angle or a contoured surface.

Screenshot 2020-10-23 100606.jpg

[[Er...]]

Plane points. Out of the blue, point groupe, maybe 10 at each taget (If cast surface) arranged equally spaced within that 3mm dia.

Analyze, correct if nessecairly.

[[Jo...]]

typically I use a point strategy looping the alignment. So that would be simply a 3 point plane. Usually the distances are referenced to the other datum features, so you would "try" A1,A2,A3 then to B1, B2, then C . Looping would start over again, by the third time those points should be on target. Some people scan the diametric target zone and try to extract the high point. IMO that's really not worth the effort and is still alignment dependent. With single points on bullseye it emulates standing on a pinhead. Our machining fixture utilize a dome so it is functionally a "point" contact.

[[Er...]]

If its a rough surface, like cast. And a big part, Its absolutely worth the effort! 100%

[[Jo...]]

Please sign in to view this quote.

ASME Y 14.5-2018 7.24.4 (A primary datum plane is established by at least three target POINTS not on a straight line)

Eric can you clarify the interpretation ? Your high point could lie within the diametric zone somewhere, but in practice if you were to simulate that with 3 post lying centered in the zone it would not be the same.. If you had 3 mini planes the diametric size, it would not level the same as 3 points on target.

[[Jo...]]

So in one case, I have a cast part with 8mm zones. The tertiary is located on a surface with casting draft. If I took the highest point it would always lie at the extreme bottom boundary of the 8mm zone because of the taper. If I iterate the alignment it will repeatably be dead centered point in the crosshair of that zone.

[[Er...]]

Yes, sure. Taking one point, in the middle of taget zone, will not give you the same alignment as measuring more points in the taget area. Pretty self explanatory. The rougher the surface, the bigger difference. I do how ever not use the highest point. I tend to use the average.

We dont measure a lot of cast surfaces, since almost all cast have pre-machines datum surfaces from the suppliers. But the few times I have dealt with this. It has Been impossible to get similar results to our (Don't know the english word, but they draw lines and paint etc to see if there is enough material etc.) without using the whole taget area. I have seen surface differences of 0.5mm within 5mm targets.

Same goes for our GOM scanner, if the target isnt a point, I dont use a point. If its square, I use a square area. Diametrical, then I do that.

I try to stick to what the drawing require as much as possible. Its many times the only thing I have that gives me a hint of the design purpouse.

[[Ke...]]

If the datum targets are identified with "X", then I use single points.

The way the datum targets are defined on your example, I would use scanned paths.
I would probably use reference points on the two sides of the part, so that I am scanning at the specified locations.
If they are on the same surface, I would recall the scanned planes into a single datum plane, for evaluation & use.

I have had good and repeatable results using this method on stampings and plastic injection molded parts. I haven't had too much exposure to cast parts that have not already been machined in areas that need to be inspected. My first instinct would be to use the same method, with a large probe tip, slow scanning, and reduce the "Inside Workpiece" outlier to 1.0?

...If the application is as John suggests (the target area where the part mounts, for machining, on a single point at each location), I would look to see what size the "dome" diameter is, and try to use the closest stylus diameter possible. I agree that trying to extract a max point on each scanned location isn't worth the time. I had repeatability issues trying this method on a flat datum area (like would be used on a fixture gage); I can only imaging that the repeatability on a cast surface would be ridiculous.

[[Ma...]]

If you use plane, then you can use circular path.
If you use point you can create cloud of points in selected diameter which is full filled - not just circle like in plain.

Recall meas. points from those 3 points and for ASME choose tangent eval.

[[Jo...]]

Good discussion. So does anyone have a reference to the "correct" ASME practice ? lol I didn't mean to imply Eric was doing it wrong, or that my method is even correct. It becomes a challenge to communicate the condition of the part form when using 3 point planes etc. , because a part could be twisted and corrupt the plane. I had a part with one Datum feature out of square, and it made all my profile and locations bad. In reality, correcting the ONE feature (datum) fixed the part. I try to include my datum geometry (location and form) in my reports for that reason.

[[Er...]]

Everywhere we have datum targets is ISO. So I havent really read that part in the asme. But I would be suprised if it is any major difference.

I dont try to correct anyone here either. Just answering Toms question.

How ever, (this is not a qoute but) if you have a diametrical target with a given size. Using one point in that target violates the ISO standard. And as I said, I dont think ISO and asme differ alot when it comes to daum targets. Except that they are drawn slightly different.

Regarding the form of the datum surfaces. I have always asumed form errors is one of the reason for targets on cast surfaces. Just to avoid things like convex planes etc.

And placing a 4 ton part on three pointy supports isnt realistic. Hence the bigger targets.

[[An...]]

Please sign in to view this quote.

See attached.

Contribution_24_10_2020.pdf

[[Br...]]

Please sign in to view this quote.

Per ASME those are datum target "areas" not points. So what you have is a planar datum feature A that is made up of those areas. A datum simulator would look something like 3X Cylinder located by their basic dimensions from one another that are 3mm in diameter. The part would then sit on those cylinders to simulate the datum.

If you are going to measure it with a CMM, I would scan each of those 3 areas and use all of the scan data as a single plane- OuterTangential fitting algorithm and appropriate filtering and outlier elimination.

[[Ma...]]

3mm circle scan strategy at those locations defined by basics to the target.

[[Er...]]

Please sign in to view this quote.

What about the surface inside your circle? I think a lot of Calypso users arn't aware of the function probabilly made for these things.


(Ignore the 3mm radius here.)

[[Ri...]]

Point Set for-the-win.

[[To...]]

Please sign in to view this quote.

If the point is on a surface with a 2° angle, do you start with a space point converted to a plane point?

[[Ri...]]

I was under the impression that there is virtually no difference between a Space Point and a Plane Point.

The only difference is the Space Point projects out radially, and the length of the projection can be changed.

I know the Plane Point is strongly suggested for 3d Best Fit Alignments, but in most inspections I've not noticed a significant difference between Space Point and Plane Point.

[[To...]]

Please sign in to view this quote.

So, in my case, I have 3 Datum Targets for A (no draft angle), with one being at a different Z height, I would create 3 points, scan them at the diameter called out in the Datum Target, use OTE and apply them to an Offset Plane.

I also have 2 Datum Targets for B and 1 Datum Target for C, which are all on draft angles, I do the same thing as above, scan at the diameter, OTE, create a line for B and the last point for C.

Does this sound like I'm on the right path?