Profile question that is driving me nuts

[wicked_tattoo]

I have been working on fixing some programs at a place i just started at. One thing that i notice in almost ever program that i pull up from the old programmer is that when he checked the profiles whether is it was form and location or form only he almost always origined to the feature and then checked it. The thing that bothers me is that he also measured 90% of these things are checked as scans and the scans were set to master. Now my understanding of master is that the first one check becomes basically the golden part and everything after is check to this first part. The other thing that bothers me is that one part that i have been working on in particular has a slot getting check for form and location and then a tighter tolerance that is checking the form only. If i check it to the datum setup for form and location and then tolerance right after the form only they fail and by the numbers it appears as though everything is getting calculated right. But in this same program (btw these are programs that have been previously written we are just trying to fix them and make them better) when the original program was written he rotated and origined to the slot and when you check the form and location it passes but when you check form only it still fails like it does when you dont align to the feature. This is driving me crazy because there is no way that it can fail form only and pass form and location..... If i need to describe this a little better i will try but that is the long version. And if i am missing something in regards to the scans let me know because i have not used them very much.

Thanks everyone for taking the time to read this, and i really hope someone might have some insight to this.

Software - PcDmis 2010mr3
Machine - Global 5x7x5
Probe - PH10M/TP200

[Addison Wood]

Could you show us the Feature Control Frames for the profile Callouts?

[wicked_tattoo]

I am not at work, but it shows a profile of surface to ABC, and then a form only check. Again this is a slot. I just dont understand that when he origins to the slot and aligns to it that he will pass form and location but when you check form only within that alignment or back to ABC that the form fails. If the form fails there is no way to be able to pass form and location. Not unless i am missing something......

[wicked_tattoo]

Any input on any of this would be greatly appreciated. I think my understanding on the scanning is correct and as far as the profile goes i just dont understand why it will pass form and location when aligned to itseld but fail form only to itself and to abc.

[bharpold]

You said the form only was a tighter tolerance, in that case it can pass form and location at a larger tolerance but fail with form only at a tighter tolerance. Are you using legacy or Xact Measure? If you rotate to the feature you will have a difference due to the fact the slot is "centered" versus being related to the rotational datum control. If you could post the profile dimensions, we may be able to see something in the defining aspect of it.

[wicked_tattoo]

My thing is though, that if you are failing form only and it has a 0.001" tolerance and then you check form and location and it has a total tolerance of 0.002" then i would think that you would still see the failure. Because if it fails form only then it should fail form and loc to unless the tolerance is crazy different.

[wicked_tattoo]

I realize that you guys cant see the print, but i am trying to say that with those tolerances the form and loc should fail to. It just doesnt make any sense.

[bharpold]

I realize that you guys cant see the print, but i am trying to say that with those tolerances the form and loc should fail to. It just doesnt make any sense.

So what are the tolerances for form only and form and location? If .001 form only checks .0015, then its out but with .002 form and location, .0015 would be ok.

[wicked_tattoo]

The form only check is .001" and form and location is 0.002"

[DungT]

I have been working on fixing some programs at a place i just started at. One thing that i notice in almost ever program that i pull up from the old programmer is that when he checked the profiles whether is it was form and location or form only he almost always origined to the feature and then checked it.

I do that sometimes, especially checking profile of a plane at a compound angle without CAD. It’s quick and easy.

The thing that bothers me is that he also measured 90% of these things are checked as scans and the scans were set to master.

That bothers me too.
To me, it is not a good idea to scan a feature for profiling without CAD, unless you know and set the correct nominal for each individual point in a scan set which is hard to do on a curve surface, compound angle…etc.

Now my understanding of master is that the first one check becomes basically the golden part and everything after is check to this first part.

Agree.

The other thing that bothers me is that one part that i have been working on in particular has a slot getting check for form and location and then a tighter tolerance that is checking the form only. If i check it to the datum setup for form and location and then tolerance right after the form only they fail and by the numbers it appears as though everything is getting calculated right. But in this same program (btw these are programs that have been previously written we are just trying to fix them and make them better) when the original program was written he rotated and origined to the slot and when you check the form and location it passes but when you check form only it still fails like it does when you dont align to the feature. This is driving me crazy because there is no way that it can fail form only and pass form and location..... If i need to describe this a little better i will try but that is the long version. And if i am missing something in regards to the scans let me know because i have not used them very much.

If it was done correctly, then it is what it is.
“Form only” checks form and orientation (angularity) of a slot in relationship to part alignment without location. Ex: if the slot is perfect (form=.0000), and the angularity of the slot is off to your part alignment that exceeds the tol limit, it would fail form only. It would pass form and location if the tolerance is large enough to absorb both angularity and location. It’s perfectly normal.

Your statement makes sense only if “form and location” tolerance is equal to or lesser than “form only”. That type of callout makes no sense, but I’m not surprised…

Anytime you check profile, the data must be taken from an alignment that was set to drawing FCF or you can check it to its theoretical location from that alignment, it makes no difference, as long as the nominal is correct. The deviation would be the differences between measured and theoretical values of each individual point within measured feature. No matter what alignment you recall, the output always the same, even if you align measured feature to itself. In other words, profile calculation is fixed immediately right after the feature is measured.

[wicked_tattoo]

So i guess i never really thought of it checking angularity.... I was concerned since when i turned on the graph it didnt really show like it was creating the form only zone equal to the points rather showing a shift. After looking at it for the last 3 hours i really have to believe that my new tolerance zone is showing it right. We have almost no tolerance when trying to hold this POS slot (0.001" form only). I really believe what i am seeing is correct. I am noticing that the size is a little small and location is a little off. It was just wierd that with the graph on it really almost looks like the tolerance zone doesnt shift but i think it is because we are looking at such a small tolerance band.

[wicked_tattoo]

I just ran another part and it still seems like it is not creating the band around the slot in a form only matter?!?! I have the settings as least squares...... I just dont know and it is driving me nuts. I mean i really think it is calculating the numbers right just maybe the graph is showing something different.

[DungT]

I just ran another part and it still seems like it is not creating the band around the slot in a form only matter?!?! I have the settings as least squares...... I just dont know and it is driving me nuts. I mean i really think it is calculating the numbers right just maybe the graph is showing something different.

The graph for form only floats around wherever the actual slot is, even if it is an inch off. The graph for “Form and Location”, is stationary. It’s fixed to theo location
If that’s what you’re asking.

[wicked_tattoo]

I really believe that the graph is representing right it is just we are looking at such small numbers that it really doesnt appear to be moving. I understand that form only creates a tolerance band around the points/scan and makes sure that they trend together and that the form and loc it creates a band around the nominal zone where the feature should be. Thanks guys i really think i am over thinking this whole thing and i think that it is calculating right. Thanks for all of the help.