【转帖】measurements on datum reference frame

yang686526

measurements on datum reference frame
hello,
if you have a part with only one datum reference frame(drf){a,b,c}. a is a planar surface which has a form control, b is a external shaft which has a position geometric control reference to datum feature a under rfs condition and c is also a external shaft which has a position control reference to datum features a & b under rfs condition. then, we put the part on the specified fixture and start building the drf{a,b,c} first on cmm and go for the measurements.
how could we measure the position controls of datum feature b & c?
(1) the position control of shaft b is to refer to datum feature a. it's just only one datume feature( plane a) in its own feature control frame, not the whole drf{a,b,c}
(2) the position control of shaft c is to refer to datum feature a & b only in its own feature control frame, not the whole drf{a,b,c}.
does any one konw this concept? it doesn't make sense for me to imagine that we could setup the drf{a,b,c} first and go for measuring the position controls of shaft b & c under this drf in cmm metrology.
many thanks,
ntuedutw
can you post a picture? i'm not sure that i understand the exact situation, but right off i don't see how you can control the position of cylindrical feature b exclusivly by planar feature a. by controlling only to a you leave 3 degrees of freedom open, you control "z"linear movement, "x" rotation and "y" rotation. this leaves "x" linear, "y" linear, and "z" rotation. cylinder c can be controlled by a and b assuming that you don't care about the orientation of the cylinder in "z". if you only care about controlling cylinder b to plane a, you would be better off using perpendicular.
i would say that calling out a positional tolerance on the datum b shaft to only the datum a plane is a drawing mistake. it should have a perpendicularity control. a positional tolerance would control the perpendicularity, but there is no locational relationship being established to any other feature at this point so position is a misnomer. a positional tolerance can be used to locate the datum c feature with only datums a,b in the feature control frame. the basic dimension establishes the location from b. it's just that at this point there is no tertiary, clocking datum so you are checking only the distance from b to c.
hi caseynick, aardvarkdw,
both you are right, it's my mistake. i think i should take a example shown on y 14.5 to describe my problem. please have your time to check out the fig 4-15(page 64) that describes the establishment of datum feature rfs.
in this picture, we could setup datum reference frame from these datum features - a, b and c. then go for measuring the four holes - dia 7.7 to 7.8. but if we want to verify the datum feature b and c that are the functional critical dimensions, how could we make it?
q:
datum feature b is only related to datum feaure a for pendicularity control. could feature b be measured under the datum reference frame establishing from datum feature a, b and c?
as same question with datum feature b, how could we measure the datum feature c?
i think it's so strage to me. we couldn't use datum feature a, b and c to setup the datum reference frame(drf) of the part, then use the drf to go back measure the datum feature b and c.
many thanks,
ntuedutw
if you make feature "b" into datum b, you cannot use datum b to control feature "b". feature "c" can be controlled by datums a and b but not c if feature "c" is to be come datum c.
if "b" is controlled only by perpendicularity to a. then the location of "b" is controlled by the other dimensions and tolerances listed on the drawing and in the tolerance block. "b" can then become datum b and control "c" relative to b and a. this would control perpendicularity to a and location relative to b but not location relative to the edges of your part for "c". so "b" and "c" float perpendicular to a and relative to each other and "b" is controlled by the other dimensions on the drawing.
again a picture would let me explain better, i think.
david
hi david,
thanks for your quick reply. i really want to attach  a picture for your infotmation but i can't make it. i don't have any web server to post the picture.
---------------------------------------------------
from tgml:
--------------------------------------------------
maybe, i could send the picture into your email box. my email: