huangyhg

datum target -- help?
asme y14.5 did not tell how to dimension the datum targets. i tried to tell people it should be dimensioned relative to each other. but they continue to dimension it from part features. would anyone know where to find such guideline except "datum targets are ... dimensioned relative to each other" - drawing requirements manual for departments of defense and commerce, 3rd edition.
thanks.
for what it's worth (little), it is also in the 5th edition of the drawing requirements manual
y14.5-1994 is vague on this question.
believe it if you need it or leave it if you dare.fff"> - robert hunter
datum targets typically describe how a fixture will contact the raw casting, stamping, or forging. that processing fixture or its equivalent inspection or assembly replicate is used to orient, align, and locate the rigid rough structure or pliable elastic structure for first cuts etc. they are also used to provide stack path links to the rf structure or define functional clamped assembly evaluation of the finished contour. the finished feature alignments, orientations and locations are dependent upon how the structure registers with those targets not visa-versa... therefore the targets should be defined in a system 3-2-1, or 4,1,1 or pliable 6,2,1 or whatever the functional or surrogate system is within its own datum structure so that it can be defined without subsequent feature process and predicted stack variation and so it can be simply used to create processing, inspection and/or assembly fixtures!!!
some designers reuse secondary or tertiary datum targets in subsequent processed operations thinking that the registry is equivalent and disregarding the surface's form effects in their stacks but that mistake can bite you if those actual errors are significant.
bottom line... define the targets amongst themselves...tolerance the processed features from them...and then "as functionally reflective" create "tolerance" other functional datum features.
paul
wow paul, my head is spinning from all that complex intermingling of vocabulary... i'm fluent in the english language and i barely understood any of that. i don't think dho's mother tongue is english so he/she is likely to skip it altogether.
were you saying that he is right and "they" are wrong? if so then i would agree that the datum targets should be relative to each other and not based on the location of any features that the part may have.

powerhound, gdtp t-0419
production supervisor
inventor 2008
mastercam x2
smartcam 11.1
ssg, u.s. army
taji, iraq oif ii
datum targets are dimensioned from each datum. you must have a minimum of 3 mutually perpendicular planes to create a datum reference frame. the datum reference frame is theoretical (doesn’t exist on part), so we simulate the datum’s from actual features of the parts.
simple example: a rectangular shaped casting.
datum targets a1, a2, a3 are all on datum –A- (shown in a view perpendicular to –A-),
they and are dimensioned from –B- & -c- (shown in views perpendicular to –B- & -c-).
note: all datum target dimensions will be basic (boxed dimensions with no tolerance).
datum targets b4 & b5 are on –B- and dimensioned from –A- & -c- as noted above.
datum target c6 is on –C- and dimensioned from –A- & -b- as noted above.
complex example: a casting with 3d contour and no flat planes.
actually the same process, except datum target points will also have a dimension from the datum they simulate.
datum targets a1, a2, a3 are all at different distances from datum –A-, the dimensions will be shown in views perpendicular to –A-, -b- and –C-.
the same process is used for b4, b5 & c6.
the datum target points are then used by manufacturing & inspection for the production of the product.
thanks to all who are helping me.
this is the part "they" assigned all datum targets. the problems here are 1) they missed z dimension for tp2a and tp1c. 2) tp1a, tp2a, tp3a and tp1c are going to "define" datum plane a. (see attachment)
1) yes. the z can be calcualted from the norminal dimensions. but it should not.
2) a physical part "sits" on the fixture will never touch those four point as the same time. only three.
comments?
i don't understand how there can be two datum a's. the datum points a1-a3 define datum a, and it is hard to imagine how it can be located .250" from itself.
if you do indeed have a planar surface for datum a, why use datum points to redefine it as shown?
believe it if you need it or leave it if you dare.fff"> - robert hunter
there is only one datum -a-. tp3a is on a stepped surface. see asme y14.5 para 4.6.3.1.
thanks.
what are "they" trying to achieve? cause this sorry, is a mess.
the example shows both tp2a and tp3a as being offset, but i'm skeptical that they could be considered as being on a stepped surface as shown in the y14.5 example.
i could be easily wrong though (and have been) as i haven't used stepped datums such as this before.
it seems to be a very awkward datum setup.
believe it if you need it or leave it if you dare.fff"> - robert hunter
a picture is worth...anyhow, here is a suggestion.
small changes:
1. all datum points are dim'ed with basic box (no tolerance).
2. added other dim's for complete definition of part.
datum changes: (there are a ton of different ways, and the product desired end result drives the requirements, but, try to keep it simple, and, more importantly, easy to produce & inspect.)
1. put a1 & a2 on same plane. put a3 on center of shaft.
3 pt's will always align.
2. put c1 @ 90 degrees to -a- on center of shaft.
reason: datum target points hold a product in relation to the datum reference frame. try to make the points as strong in the degree/vector of the datum as possible, i.e.; the example shows a2, a3 & c1 at less than 45 degree's to each other on two different surfaces. the odds are that those three points will not all make contact because we are already resting on a1 (4 points strong in the -a- vector). c1 is not strong in the -c- vector, so it will not control -c- location as strong as we would like.

__________________
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]