True Position Datums

[Bob Pierce]

I have little to no experience using true position. The concept I need to grasp is this. Does the part alignment datums have to be the same as the datums specified in the feature control frame or can these be defined in the dimensioning window. I have a print with several true position callouts using different datums. Must I do a new part alignment to each call out prior to creating the dimension?
Also can I do a true position of a feature set?
Thanks to all

[Matthew D. Hoedeman]

I have little to no experience using true position. The concept I need to grasp is this. Does the part alignment datums have to be the same as the datums specified in the feature control frame or can these be defined in the dimensioning window. I have a print with several true position callouts using different datums. Must I do a new part alignment to each call out prior to creating the dimension?
Also can I do a true position of a feature set?
Thanks to all

You DON'T have to make an alignment for each different FCF, you can do that with datums in the TP dimension window. HOWEVER, I would do so because Pcdmis WILL shift the part around as if it were sitting on a hard gage to get you the best possible deviaiton values for the feature you are dimensioning. This can give you perfect zero deviations and we all know that no one understands HOW your part can check perfect, time after time after time. By making alignments to match the FCF, then all you need to is dimension the TP and it will show you actual deviations. Also, you MAY want to use PERP TO CENTER LINE when doing TP, this will eliminate the surface deviation. AND, you SHOULD take at least 1 surface sample to control your depth of touches for the hole you are measuring.

[Wes Cisco]

With the new forhead super tp features the following may not apply, but for 3.7MR3 and earlier the answer is Yes you must create a new alignment prior to dimensioning for each and every different FCF callout. So if you have prints like some of mine with 30 plus datums and blunderneers who tolerance side by side features to A/B/C, C/B/A, & B/A/C then you will have many many alignments. If the new 4.x tp allows otherwise I trust that James or someone will provide you with some guidance. HTH

[Bob Pierce]

true position datums

If changing the alignments to the FCFs ( for dimensioning only ) and I am using a CAD model , do I need to do origin offsets each time?

[kbotta]

With the new forhead super tp features.....

You almost sound like Jan! Where is he???

[Matthew D. Hoedeman]

You will IF you have to report CAD POSITION values. However, you COULD do iterative alignments (possibly) for each one. HOWEVER, I would STRONGLY suggest that all the extra alignments get done at the END of the program, then each dimension directly after it's alignment. ie Do the alignment, do the TP's (all of them) for that alignment, then the next, and so on. This will help you keep from screwing up your program.

[inspector212]

Fit to Datums

Fit to Datums does not work correctly. Don't use it. I had a part where the feature I was inspecting was more than 5mm deviated from nominal. When using Fit to Datums PD-DMIS shifted B and C 5mm to bring the feature into spec. I reported it to B&S and they acknowledged the bug and have been working on it. FYI V4.1

[Bob Pierce]

true position datums

Won't I be required to specify datums in the dimensioning window or are you suggesting I use the legacy dimensions?

[inspector212]

You can use the FCF's or the legacy but just don't use the Fit to Datums button in V4.1.

[rangerboat72]

You almost sound like Jan! Where is he???

Where is he, I thought Jan was a she ? ? ?

[Jan d.]

You almost sound like Jan! Where is he???

He's been hiding, OK, cat's out of the bag!

I just finished a week of PC-DMIS advanced training.


In 4.x you do NOT need to do real close alignments at all! You just take your points, declare your datums and evaluate against your datums. Super simple. But be careful: the closer your tolerance, obviously the better your initial alignment needs to be.

In NC, which is what I use, you can not use alignments to finetune your machine motion. So after making your CAD the same as the part (which is the initial alignment), you measure all points.

If you only need to evaluate your FCF, just do that right after all motion stops, and print them. Done.

However, as guru-Kevin pointed out to me, some people do not just want to see the FCF evaluated, they also want to see where the feature is with regard to the basic dimensions. To me, this is totally irrelevant (how can you tolerance a basic dimension?) and a waste of time (all you have to do is evaluate the FCF, why look at where the basics are???), however, I was reasoned to the dark side.

Therefore, after all motion stops, start to do alignments to the actuals so your basics look OK and then evaluate and print your FCF's.



Jan.

[kbotta]

Cool.

[george frick]

You almost sound like Jan! Where is he???

he? i've been thinking all this time "jan" was a slick chick- a she.

[JamesMannes]

he? i've been thinking all this time "jan" was a slick chick- a she.

It's one of those androgenous names. Who knows Jan could be a chick with a !#%@. (Just teasing ya Jan).

[george frick]

It's one of those androgenous names. Who knows Jan could be a chick with a !#%@. (Just teasing ya Jan).

or a "nappy-headed ho"!