Bogus Datum scheme?

[John Riggins]

I have had a problem getting an iterative alignment to work with a part (see attached).
This part has nothing square to anything.
It is a "V" shape.
Our customer put 2 of the -A- datum points on one leg of the V and the third -A- point on the other leg (leg 2).
Then they put the 2 -B- points on leg 2.
The program runs offline, but when I try to run it online I get an Iterative alignment error.
I substituted the -A2- point with a point on leg 1.
Program runs ok.....
Is the datum scheme bogus or am I a bozo?

[jon_dav_uk]

For an iterative alignment to work the vectors of the 3 datums need to be as close to 90 degrees from eachother as possible (i.e. vectors 1,0,0 - 0,1,0 - 0,0,1) or else u will get the error u r seeing. They do work if the vectors r close to this but not always

You could try a 3D best fit alignment

Jon

[John Riggins]

Thanks Jon,
I did try a bestfit, but this part is a bit twisted so it made things worse.
What I really need to know is if the datum scheme is legitimate.

[jon_dav_uk]

It does look a bit strange but then again the part has no geometric features and is completely freeform so maybe not. Is point A2 on the same face as the B datum points or the A datum points, hard to see from the pic as not sure if the blue line is a leader pointing to A face or just another feature if u get my drift. If its on the B side this point is probably throwing the iterative more than any other and this point seems to be the oddball. If it is on the B side of the part try taking it on the same side as the A points and see if the iterative works then,not conforming to drawing but proves a point

[jon_dav_uk]

lol just seem the substitute A2, have u tried using that ?

[John Riggins]

Jon,
Yes, A2 is on the same face as the B points.
And yes, I did take the A2 hit on the same face as the other 2 A points and the Iterative alignment worked.
I would like to complain to our customer that they have a bogus datum scheme and get them to change it to what I have done to get the part measured.

[jon_dav_uk]

Yes I would agree totally, the substitute datum is the correct one,not sure why they picked the other one as it makes no sense, glad you got it sorted just remember....the customer is usually wrong

Jon

[Rick_Stanich]

Looks like a "snap" error on the drafters end, snapped the wrong end of a line?
I concur with swapping surfaces fo A2.

[Peter Warcholyk]

Ditto,

Datum "A2" should be on the same face as "A1" and "A3".

ZydecoPete