【转帖】datum definition, pattern of holes

yang686526

datum definition, pattern of holes
quick question for you gd&t gurus out there.
i am working on a part that has two dowel pin holes used for alignment of a part.
i think it is possible to define the two holes as a pattern and then define the pattern as a datum to define secondary and tertiary datums to be used for location of other features.
so the datum axis is centered between the two holes (according to asme y14.5-1994 section 4.5.8), secondary datum coincides with the axes of each hole, and tertiary intersects secondary at the datum axis and is perpendicular to primary and secondary.
i have not done something like this before but it seems reasonable and mirrors part function. anyone see any problems with this?
will post image/link asap for clarity.
thanks!
joebk
check out our whitepaper library.
are the holes coaxial?
link is below. to see drawings you will have to follow the second link on the main page. for some reason the tgml tags barf when i add the thread number as part of the link. hope this helps!
the picture comes up a little fuzzy, and i can't tell which datum is which or how they are controlled.  the best i can recommend is to follow the standard example a little more closely, and establish your primary datum (datum a) as the face of the plate, then you could use the pattern as your secondary datum (datum b), controlled relative to a (unless the pin perpendicularity does not need to be controlled).  i am sure that others here will join in helping clarify this matter.
sorry, googlepages can be a pain sometimes. you should be able to click on the jpg image of the drawing to view it full size.
there is also a high quality pdf you can view/download that can be viewed, panned, and zoomed easily.
hope this helps!
yes, you can do it the way you show in the drawing, but i would not call out datum d the way you did (hanging off the tolerance block).  identify each of the pins seperately as datum targets, e.g. d/1 and d/2.
sorry, i didn't see the pdf link before.  i agree that your method is legitimate.
shouldn't the pattern be controlled to the b datum? what is keeping the entire pattern in callout 2 from pivoting around the d datum hole? i don't think that specifying the d datum the way you did is legit. you can specify 1 hole as a datum but when you put 2x with the hole size, that changes things. if your intent is to create a datum plane between the two holes then each should be it's own datum and whatever you are trying to control to it should be called out with respect to d-e if the other hole is datum e.
  you have the part fully constrained in all degrees of freedom. why aren't you just calling out the features and patterns wrt a, b, and c? it appears that this will accomplish the same objective without the extra datum.
powerhound
production supervisor
inventor 11
mastercam x
smartcam 11.1
ssg, u.s. army
taji, iraq oif ii
the datum d callout is attached to the fcb, not the hole.  there are two holes covered by that fcb.  those two holes therefore define the datum, not one hole.  if it were one, how would you choose which one?
duhhhhhh, never mind. i guess i should stop being in such a hurry when i read through these things. i just read the op again and actually checked the reference in the standard and i see the real deal now. sorry about that.
powerhound
production supervisor
inventor 11
mastercam x
smartcam 11.1
ssg, u.s. army
taji, iraq oif ii
i know the feeling... i've been guilty of that more than once.