【转帖】profile tolerance - 2 different coord systems

yang686526

profile tolerance - 2 different coord systems
-lofted part with datums defined using 6 datum points on surfaces.
-mold for part, with convenient surfaces for datums.
since both part and mold are defined relative to three primary datums, is it a matter of concern that the datums are not defined exactly the same?  does it matter that the mold has datums defined from flat orthogonal surfaces and the part from six datum points?
i think i know the answer, but need to explain my reasoning to the project manager, and don't want to seem too ignorant on the issue (especially if i overlooked something).  thanks!
check out our whitepaper library.
ewh,
   how easy is it to design a molded part with datum targets such that you can use the datum targets on the female mold?  the really serious gd&t types would probably insist on you doing this, but at some point, it is not possible.
   i believe the crude assumption with tooling is that you need to be ten times as accurate as your part.  if your mold is fabricated with that accuracy, then your part datums can be located from your mold datums.  
                      jhg
thanks for your response, jhg.
my reasoning is that the part and the mold are each tied down to three primary datums, regardless of where those datums are.
as you state, mold tolerances are much tighter than the part tolerances, and it only seems reasonable to use the most efficient method to inspect the items.  however, to please the customer, we have to use their part datums when inspecting the part.  there are situations where the part inspection is "bought off" the mold.
since the mold tolerance is always tighter than the part tolerance, wouldn't the part always be within it's own tolerance, even though the mold uses 3 different datums?
what is the worse case situation using two different datum systems?
ewh,
   when you translate from one coordinate system to another, you lose accuracy.  if you lose enough accuracy, your part goes out of specification.  it is desirable to use the same datum system throughout your manufacturing process.  then again, if your losses are tiny enough, they are irrelevant.
   some people have trouble with terminology like desirable and preferred.  they insist on absolutes.  
   is your customer inspecting your tooling or auditing your drawings?  if your parts pass inspection, they should not care how you do it.  this sounds like the equivalent of somebody specifying the drill size for tapped holes.
                   jhg
ewh,
when datum target points are used to constrain all six degrees-of-freedom for a component鈥檚 drf the dimensions describing their location have to be basic. typically those basics are relative only within the system of targets themselves so that the 鈥渘est鈥?definition can be transported simply and identically to other details such as machine registry fixtures, gages, or even to iterative probing cnc or cmm programs that hunt for the actual surface condition that only an actual physical registry would provide.
using the same targets in two or more different systems will (theoretically) never deliver equitable registry with the one system! that said 鈥渨hat鈥檚 the marginal difference鈥濃