Reporting Profile on Inspection Report

[MachineShopGuy]

What is the correct method to report profile on an inspection report(assuming unilateral tol.)? Our CMM outputs worst case Min/Max deviation, and while we recognize that Profile is 2X worst case, it seems that providing actual deviation, and direction would have more value to our customer. Is there a standard way to report?
Bob

 

[KENAT]

thread174-227626

KENAT,

Have you reminded yourself of faq731-376 recently?

 

[KENAT]

For ASME I'd think reporting max deviation was probably appropriate so long as that is in tolerance. If you are out of tolerance then direction may help in dispositioning the part.

For ISO may be different.

Could you clarify what you mean by "we recognize that Profile is 2X worst case" I may be misunderstanding what you mean.

KENAT,

Have you reminded yourself of faq731-376 recently?

 

[MachineShopGuy]

By 2X worst case, I am refering to the fact that profile would be controled by 2 equidistant boundries on either side of basic profile...ie a profile of .010" (again assuming unilaterial tol. zone) must fall within a zone that is +/-.005 off basic. ie.. if feature was at worst case of +.004 (incorperating size, form and location), it would require a profile of .008 to be acceptable. So... Do we report Profile as .008" or +.004"?

 

[KENAT]

That's what I thought you were saying and I wasn't following why.

I'd think you'd report it as +.004. Just because at the worst point it's .004 out on the + side doesn't mean it's .004 out on the - side, so saying you've achieved .008 profile doesn't sound correct to me. You may be holding .004 profile, just not bilateral.

The 2 equidistant boundaries are defining the min & max. I dont' think it makes sense to try and report as achieved profile.

If it was just a +-.005 dimension, and you measure +.004 would you report is as having achieved +-.004?

KENAT,

Have you reminded yourself of faq731-376 recently?

 

[MachineShopGuy]

OK...Its late, and as I read this, I realize a should have stated assuming Bi-Laterial Tol. per my example. Does this change how you would report?

 

[KENAT]

No, you did state bi-lateral somewhere. Maybe I'm the one not understanding your point.

.010 bilateral with equal distribution is simplistically like +-.005 as you say.

If you measured +.004 on a conventional +- tol would you report is as having achieved +-.004, I'd assume not.

So I can't see why you'd do equivalent with a profile.

If you measure +.004 as your worst case on the + side, this doesn't mean you've got -.004 on the - side so I don't see that you can meaningfully say you've hit profile of .008.

While the requirement may be .010 bilateral equal distribution, you may have achieved .004 unilateral or some unequal bilateral.

Maybe I'm out of my depth and should leave it to someone else.

KENAT,

Have you reminded yourself of faq731-376 recently?

 

[PaulJackson]

MachineShopGuy,

Doubling the worst deviation from basic profile only works when the profile is specified equal-bilateral! When the profile is described via illustration of the tolerance zone as either unilateral (all more material or all less material from the basic profile) or unequal-bilateral (unequal specified amounts of more and less material from the basic profile) then... doubling the worst deviation does not work!!!

Adopt a policy of reporting the deviations as "0.008 +/- 0.004 is +/-X.XXX to +/-X.XXX where one is the more material and the other is less material"... When capability predictions are required use both specified + and - constant limits i.e. EQUAL BILATERAL +0.004, -0.004... or UNILATERAL +0.008, -0.000 or +0.000, -0.008... or UNEQUAL BILATERAL +X.XXX, -X.XXX (given that the ABS of the sum of the two equals 0.008).

If the policy is adopted... design, manufacturing and quality can all interpret the measurment results unilaterally, equivalently, unanimously, or harmoniouusly or however all disciplines percieve them according to a written procedure.

Paul

 

[KENAT]

Paul, OP was talking about equal-bilateral.

If I've understood you correctly you agree that reporting measured tol variation in profile form doesn't make sense though, correct?

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at

http://eng-tips.com/market.cfm?

 

[PaulJackson]

Kenat,

I believe that the best way to report profile is to report the extreme + (more material) and - (less material) deviations from the basic profile.

I used to think that it was OK to just double the worst deviation from an equal-bilaterally specified profile tolerance but... reporting a deviation of +.002 to +.003 in comparison to a deviation of -.003 to +.0025... as .006 in relation to a .008 specified profile tolerance... lacks meaningful characterization of the surface condition for inspection scrutiny. One surface has a maximum measured variation of .001 but averages +.0025 more material while the other has a maximum measured variation of .0055 and averages .0005 less material.

If the tolerance is specified .008 unilaterally (0 to +.008 more material from basic profile) how is one to report a negative result? The same goes for unequal bilateral if the tolerance is specified (-0.002 to +.006 from basic profile) how would one report -.003 to -.001 measured results?

I think that if people just agree to report a profile tolerance of .008 (-X.XXX to +X.XXX) as specified equal bilateral, unequal bilateral, or unilateral... measures X.XXX "less material" to X.XXX "more material" then you will have one common way of reporting measured results for all profile specifications.

Paul

 

[KENAT]

I think that's what I was trying to say.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at

http://eng-tips.com/market.cfm?