huangyhg

profile tolerances and datum references
this question may be a bit hard to explain but hear goes.
we have a flat part with a hole on one side - this side of the part is squared off (a 1" x 3" x .25" plate). on the other end of the part a series of arcs/radii combine to describe a "cam" shape if you will.
all of the basics that describe the theoretical perfectly shaped "cam" profile come off the centerline of the hole. however the profile tolerance on the "cam" shaped surface references one of the outside tangent points (on one of far radiuses) as the tertiery datum. essentially, lay the part flat (datum -a-), two points on the bottom edge (datum - and bump the tertiery datum, -c- against a flat surface for the third datum. this is how the current feature control frame calls out the profile tolerance.
my question is this; since all of the basic dimensions that define the profile come off the centerline of the hole, should the tertiery datum be the that centerline? does it matter if the product function dictates that the end of the part (a part of the profile) the controlling point? is there any part of the ansi spec that requires/recommends that the referenced datums be the same point/plane from which basic dimensions are originated?
just curious what others think here.
i think it makes more sense that the tertiery datum be the centerline of the hole - it just seems more logical - but i know from a product function standpoint it doesn't make any difference.
thanks!
it's always more clear to me to see a pic.
who designed the part, and why did that person choose the datums?
mtncrawler,
never apply datums to imaginary features like centrelines. apply them only to real features.
your hole is a legitimate datum feature as per asme y14.5m-1994. the face, the hole and one edge locate your part completely as required by the standard, as does the face and two edges. if i used the hole as a datum, it would be the secondary datum, with the edge controlling only rotation.
the real question is how you ought to fixture your part for manufacturing and inspection. the manufacturers and inspectors are supposed to use your datums for this. this ought to affect your choice of datums.
good drafting practise is to run dimensions from the datums. this is not always convenient, and alternate dimensioning procedures sometimes show relationships you are interested in. then again, if you are interested in these relationships, maybe you should apply datums accordingly.
jhg

if my space vision is still intact there should be an angle in the mix to establish the beginning and/or the end of the radius location. that would be the tertiary datum even if you resorted to a datum target with tolerance.
not exactly the same but you get the idea
mtncrawler,
you had stated 'flat part as primary datum feature, edge as secondary feature'. i see no problem with that up to that point. however i would think that the hole would serve as the tertiary in preference to the tan point.
i believe this would provide a functional reference frame for the fab and inspection of the part as you have explained. (this is of course without the benefit of seeing the drawing.)
thanks for the feedback all. i know its hard to visualize without a pic.
drawoh, yes, i meant the hole as a feature not just the centerline. we are currently inspecting/nesting per the datum -c- as is. however, our inspector was using the hole - because that was the origin for all the basics. essentially, it amounted to a shift in that axis; nesting to the right against the current specified datum .vs. a shift left, locating the hole.
hence my question, even though its not good drafting practice, the specified datum should be used or, if in agreement with part function, use the hole as the datum and all the basics would make good sense coming off the datum.
as it turned out, we'll change the drawing so that the basics originate from the current datum not the hole.
thanks again all!
mathematically, it doesn't matter where your dimensions come from because there is no accumulation of tolerance on basic (theoretically perfect) dimensions. convention, however is to make it easier for the programmers, machinists and inspectors by dimensioning from the datum.
which you select for the datum feature is dependant on the function. which feature controls the part, and which feature is more relevant to the position of the other features; the hole of the tangent of the arc?
using the tangent of the arc isn't an issue because only one point of contact is needed to establish the tertiary datum. that point of contact is against a plate simulating the datum, which is mutually perpendicular to the primary and secondary datum simulators.
the reference that thundair uses from the tec-ease site is relevant, but with some consideration. in that reference, the tertiary datum feature is a planar feature specifically set at a basic angle to the other datum features, whereas here the datum feature is an arc. either way, the point of the tec-ease graphic was to show that the tertiary datum (regardless of actual datum feature orientation to the primary and secondary datums) is always mutually perpendicular to the primary and secondary datums.
i hope that helped rather than confused!
jim sykes, p.eng, gdtp-s
profile services
cad-documentation-gd&t-product development

__________________
借用达朗贝尔的名言：前进吧，你会得到信心!
[url="http://www.dimcax.com"]几何尺寸与公差标准[/url]