DTM B True Pos. to DTM A

[Farmer]

Good Morning Learned Peers. Hopefully one of you could answer (What I hope is) a simple question. The part that I am currently programming for has a a true position call out for Datum B (Ø 0.90) to Datum A (Please see att). The JPG that I have attached is the plan view. The underside of this piece is datum A (This face is on the cmm bed). Im guessing that this tolerance is something to do with perpendicularity as obviously there can be no dimensions from B to A. How would I tolerance this? I am measuring the cmm bed as my DTM A.


TIA





F.

[rangerboat72]

If you are required to measure it, you should measure it as a cylinder and report the perpendicularity to your datum plane. Is the material to thin to measure? Sometimes we just overlook that requirement on sheet metal parts. It's just a general tolerance that applies to your secondary datum due to the fact that your primary should generally be flat, your secondary should be perpendicular to your primary and your tertiary should be positioned to your secondary (and perpendicular to your primary). If it is thick enough, measure it. If not, don't sweat it.

[inspector212]

If you are required to measure it, you should measure it as a cylinder and report the perpendicularity to your datum plane. Is the material to thin to measure? Sometimes we just overlook that requirement on sheet metal parts. It's just a general tolerance that applies to your secondary datum due to the fact that your primary should generally be flat, your secondary should be perpendicular to your primary and your tertiary should be positioned to your secondary (and perpendicular to your primary). If it is thick enough, measure it. If not, don't sweat it.

I concure.

[Underspec]

I would measure datum -A- and not the cmm bed if possible. Hold the part differently?

[Paul Sarrach]

Yes it is a perpendicularity call out. Its one of the many stupid call outs that we see. Just remember if you do do this call out, check only the pierced and not the breakage.

[rangerboat72]

I would measure datum -A- and not the cmm bed if possible. Hold the part differently?

I meant to mention that also. Good catch!

[inspector212]

The cmm bed is a good datum simulator. There is nothing wrong with it.

[Underspec]

The cmm bed is a good datum simulator. There is nothing wrong with it.

If the part is bent a little and your tolerance small, then yes it would be a problem. It's a bad habit to use the cmm for any datum. Set it up to use A90B0 and A90B180 if possible.

[rangerboat72]

The cmm bed is a good datum simulator. There is nothing wrong with it.

If there's a perpendicualrity callout (position to [-A-]), there's most likely a flatness requirement that may need to be measured also. That would be my main concern with using the CMM bed as a datum simulator.

[Paul Sarrach]

Just another thing. Are you checking this free state or on a fixture? Most sheetmetal parts are required by the print to be checked in the restaint condition, there for you really can not check it. I have been in SheetMetal for almost 20 years, I always just make the PPAP Report good. They are just wanting the hole pierced perpendical to that surface, and thats thier way of stateing it.

[inspector212]

If the part is bent a little and your tolerance small, then yes it would be a problem. It's a bad habit to use the cmm for any datum. Set it up to use A90B0 and A90B180 if possible.

If there's a perpendicualrity callout (position to [-A-]), there's most likely a flatness requirement that may need to be measured also. That would be my main concern with using the CMM bed as a datum simulator.

What would be the difference with a best fit plane constructed from points and the cmm bed? None.

There should be a form control on A, like flatness.

[#2#]

What would be the difference with a best fit plane constructed from points and the cmm bed? None.

There should be a form control on A, like flatness.

There would be a difference. Say the plane is curved in one dimension. Lets say the flatness is .010. Placing the part on the cmm bed will hit only the high points and your plane will locate there. Taking points on the surface will hit both high and low points and a best fit plane will be an average of the points. So, the best fit plane would most likely be around .005 up from the cmm bed.

[mdobb829]

If there's a perpendicularity call out (position to [-A-]), there's most likely a flatness requirement that may need to be measured also. That would be my main concern with using the CMM bed as a datum simulator.

I agree.....I am assuming there is another view, since Datum [-A-] is not shown in this view. The true position call out shown is an incorrect FCF. If you were checking true position of this hole you would need a datum reference frame that eliminates the 6 degrees of freedom. What is implied by this type of call out, is the perpendicularity of the cylinder created from Datum [-B-] to Datum [-A-] not the true position. If there is no flatness call out for Datum [-A-] using the CMM bed is acceptable, but it is always best to measure the feature itself. I would orientate the part so that all the features can be measured in the same set-up.

[Underspec]

Goodluck is right, if you're using the cmm bed you're just using the high points on the datum. If it is sheetmetal or plastics then usually you want it in a restrained condition if you plan on using the cmm anyway.

[dwade]

Measure datum A then measure the hole a cylinder providing there is length enough to do so. Intersect the cyclinder with the plane, origin to that point, then tolerance both ends of the cylinder. One SHOULD report near 0/0 the other will show some deviation.

My .02.

Duane