DCC Alignment - Is this bad practice?

[[email protected]]

...To use DCC alignment, then re-use the DCC alignment features to measure from.
That is, when the alignment is applied, do those DCC features get moved, or those are all still at zero?
In this case I have a plane for Z=0, a line at X=0, and a line at Y=0.

I thought I should be able to measure accurately from them.

Or is it better to do the DCC alignment, then create separate features from which to measure?


Thank you,
J

[Don Ruggieri]

When you say "from which to measure", I am thinking maybe you mean "from which to dimension"?

There is nothing wrong with using alignment features in constructions, dimensions, etc. assuming you've programmed in sufficient hit patterns to accurately represent the feature in question. For example, if you have a 75mm diameter hole, and you've only taken 4 hits in it for an origin in an alignment, I would suggest that is not sufficient for dimensioning roundness.

[[email protected]]

In the sense of PC-DMIS, "dimension" is the correct term.

I have since discovered that, although the part is straight, with perpendicular sides, the plastic has very slight "sinks" that affect the construction of the features. Hitting all the level points yields the desired dimensions. Hitting the sunken areas causes it to come up short. If I was measuring with calipers, I'd be hitting the level spots, so that is how I'm locating my hit points now (basically ignoring the sinks). For the purposes of fit, this is acceptable.

My hit points for my DCC alignment were on some of the sinks, which is what was messing things up.

Thanks,
J

[John Riggins]

For datum features to create the Final DCC alignment (the one thats used for dimensioning) avoid the sinks.

[Kevo_cmm]

When you say "from which to measure", I am thinking maybe you mean "from which to dimension"?

There is nothing wrong with using alignment features in constructions, dimensions, etc. assuming you've programmed in sufficient hit patterns to accurately represent the feature in question. For example, if you have a 75mm diameter hole, and you've only taken 4 hits in it for an origin in an alignment, I would suggest that is not sufficient for dimensioning roundness.

+1.
Almost NEVER is it a good idea to use the minimum required points for datum's or features!
Lots of extra points dont really add very many extra seconds to a feature measurement or minutes to a part program, and provide much better repeatability and believeability.

Something else you can do with a diameter measurement is to use maximum inscribed for id's or Minimum circumscribed for od's. That will discard the points in the sink areas from circle construction. Doesn't work well for partial circles though.

[william]

On the first part I programmed I had an issue when I used some of my original alignment features as true positioning datums. I used my level plane and my xy origin feature (which was a circle and defined as datum D on the drawing). My program seemed to run well for a few parts but after a couple runs of the program on several parts it started to report true positioned features as being off location be more than a hundred inches. (not 100 thou) I still don't know enough to say that perhaps there wasn't another issue that was causing it but what I do know is that as soon as I re-measured the features necessary to take my defined datums outside of my alignment that problem went away and hasn't been back since in two production runs of the same parts. Anyone have some insight on that?