yang686526

gd&t gurus
this might seem like a simple question but i asked some gd&t experts and i got different answers.
suppose you have a ring with two (cylindrical) surfaces. there are two surface profile on each surface with the tolerance .06
the way the ring is dimensioned is that the basic radius for the inner surface (call it surface u) and the basic dimension for the thickness t are given. this means that the radius of the outer surface (call it surface p) is inferred from the radius of u and thickness t.
now does the profile on surface t establishes the tolerance for t or radius of surface p?
stated differntly, which combination of the tolerances do we have:
a: surface u +/-.03
thiscness t +/- .03
surcace p +/- .06

or

b: surface u +/-.03
thiscness t +/- .06
surcace p +/- .03
check out our whitepaper library.
how can t be both a thickness and a surface?
t is the thickness. i aplogize for the type the correct sentence should be:
now does the profile on surface p establishes the tolerance for t or tolerance for radius of surface p?
neither a nor b. if "t" is a basic dimension, it has no tolerance. surfaces u and p each have a profile tolerance of 0.06 relative to their respective basic dimensions.
suppose that basic dimensions, which are only given for surface u and thickness t is 14 and 1 respectively. now suppose you inspect a part that has u of 14.02, t of .96 and therefore a p of 14.8 do you reject this part or accept it? what i meant by the tolerance for t in this case is:
1-.96 = .04
sorry for another typo, u of 14.02 and t of .96 makes p of 14.98 and not 14.8
i agree with btrueblood. surface p would have an inferred basic dimension of 15 with a profile tolerance of .06 (14.97/15.03).
answer: neither a nor b. the thickness actually has no tolerance associated with it.
basic dimensions are added with no tolerance, always. they establish a virtual ideal condition to which actual conditions are compared.
the profile tolerance on surface "u" does not affect any tolerance frame placed on outer surface "p", unless surface "u" as used as a datum reference to measure "p".
that's the core of gdt. not just dimensions and tolerances, but the references used to measure.
if outer surface "p" has a profile tolerance that does not relate to inside surface "u", it is possible for "p" to nearly intersect "u" (zero thickness) and still be in spec as long as its radius is correct. there is nothing relating the position of "p" to the position of "u".
if a single profile callout has two arrows pointing to both surfaces, the profiles are measured simultaneously from the virtual condition. this is better, but does not guarantee thickness or concentricity.
chances are, you want to measure profile of outer surface "p" with respect to "u". you may also wish to spec runout of "p" w.r.t. "u". this certainly demands clarification from whoever originated the drawing.
alithepro
my question to you:
what are the datum references for each of the profile callouts?fff">
i could be the world's greatest underachiever, if i could just learn to apply myself.
the datum references for the profile is a,b,c were a is the side surface of the ring (the surface where the ring lies when you leave it on a table), b is center line of the ring, and c is a point on the ring used for clocking control. by the way this is not realy a ring. it is a gas turbine disc. the point to re
it is not correct for b to be the centerline of the ring without relating it to a surface or feature. what would establish it for inspection purposes? in this case, it needs to be the centerline of either surface p or surface u. it is incorrect to label a centerline as a datum, but it could be labeled "cl a" (centerline of datum a).
i realize that this may be beside the point, as the drawing probably already exists, but to conform to y14.5, it needs to be corrected.
ewhfff"> is correct. the gdt scheme is not valid. "b" must be the centerline of a physical feature, not a virtual wishful non-feature.
the fact is that your customer has put you in the position of interpreting incomplete and incorrect information. anyone's best guess is likely to be wrong, as there is no true correct answer based on this information.