【转帖】question regarding setting up datum planes

yang686526

question regarding setting up datum planes
please take a look at the attached image and let me know if i can define the flat portion of the semi-sphere as my datum -a- and still measure the position of rectangular cut-out.
thank you,
adrian
the flat portion is a good datum a, and the perpendicularity of the rectangle can be controlled from it, but the sphere diameter will be needed as a datum to locate the rectangle's position (or perhaps vise versa, if the rectangle is important than the sphere diameter position relative to the rectangle).
the spherical portion of the part is inadequate for a datum, as any datum radius must have more than 180° to be used as a datum. it has to be captured to adequately nail down the location the datum feature. (i hope this is making sense--it's a lot clearer in my head!)
i would use the feature itself as datums b & c. the length could be datum b, the width datum c. since this is a critical feature, this orients everything to that feature. in essence it is the first feature from which everything proceeds.
thanks both of you!
here comes the dumb question, what happens if i specify datums a, b and c such that the rectangular cut-out becomes not accessible when the part is clamped on a cmm table, is it ok? when we define datum planes, do we think of the cmm process or is it mostly related to critical features on part.
i am asking all this because i am new to gd&t and trying to make sense out of it all.  
clamping the flat on the cmm table does present a problem.
i see the validity in wgshere's point regarding the sphere, but there are other ways.
the od of the flat portion could be datum b, and also three target datums, 120° apart could be defined on the spherical diameter for machine setup, and subsequent inspection.
i believe 2 datum features will suffice for your needs. the flat surface as primary and the spherical dia as secondary.  it appears that you are still using the 1982 version of y14.5.  is that true?
no, i am using the 1994 version, but like i said..i am just getting my feet wet with this.
my dilemma was with undercut features when part is clamped on a cmm table for inspection. the way i understand it noe is that datum planes should be identified such that cmm operator can have access to all the features. correct?
the datum identifier for a rather threw me on your attached image.  you need not have the flat surface down when using the cmm.,
the most important consideration in selecting the datum features is function!
which features on the product constrain its translational and rotational degrees-of-freedom as it functions or as it is functionally assembled with other parts? decide what locates and orients what?
selection of the primary... from your sketch
if the sphere... it is capable of stopping 3 translational and 0 rotational degrees-of-freedom.
if the plane... it is capable of stopping 1 translational and 2 rotational degrees-of-freedom.
if the opposing slab surfaces... they are capable of stopping 1 translation and 2 rotational degrees-of-freedom.
if both opposing slab surfaces simultaneously (the internal cuboid)... it is capable of stopping all but one translational degree-of-freedom (along its axis or the intersection of its median planes).
the secondary can only constrain degrees-of-freedom that are not already constrained by the primary according to its ability and the tertiary subsequently likewise.
identify which feature locations and orientations are dependent on which according to function and that will help to determine which is primary, secondary, and tertiary.
  
i disagree with the comment
quote:
the spherical portion of the part is inadequate for a datum, as any datum radius must have more than 180° to be used as a datum.
... the upper swash plate of an ultra-precise spindle on a circularity (roundness) measurement device typically has a 10%-15% spherical section providing a focal point above the plate's surface that serves as its functional primary datum while a pair of tensioned jack-screws provide adjustment for two of the unconstrained rotations while the third remains unconstrained... the air bearing spindle.
inspect-ability and substitute functional process registry features considerations can supplant the selection functional datum feature but they should only be employed if functional datum features cannot be accessed or reasonably inspected.
paul   
   
using a cmm, you can establish a datum plane (or planes) from any surface, given enough points.  so, you can use a datum simulator plate offset & parallel from the cmm bed, then probe off of that plate's underside to establish your primary datum plane.  then, engage the primary datum feature (-a-) to the underside of the simulator plate and securely affix your workpiece in space (do not affix it to the simulator plate).  then, remove your datum simulator plate, and go about your inspection.
for centering based on the outside diameter of the spherical surface, you can use a smallest-diameter cylinder to simulate the effect, or use centering vices or equivalent for a convenient setup.
one of the major shortcomings that i see in cmm use is that inspectors frequently forget the basics of fixturing that they used before cmms.
using cmm contact points directly from the surface of the workpiece introduces errors; whether or not the errors are significant is dependant on the surface quality and the tolerances.
jim sykes, p.eng, gdtp-s
i firmly agree with paul.
it might make a difference whether the function of this part is inside a precise mechanism...or to be used as a boat anchor.
if we do not check according to function:
cmm checks may not correlate to the build!
remember that if they don't correlate...the cmm is nothing more that an expensive toy.
michael