【转帖】is parrallel datums required

yang686526

is parrallel datums required?
greetings
i am not an"expert" at gd&t however i know what i know and it is based on very good training and many years of machining expereince and then design experience.  i was confronted by an "old" = "expereinced" checker this week making suggestions for improvments to an initial rough draft of a model i was detailing.  this model is of a cast stainless part cast into an aluminum cylinder housing.  i espablished datum a in the bore axis, i established b on the surface mating to the crank case, and i established c on the surface that the support post is mounting to. (horizontal engine cylinder)  b is perp to axis a, and c is parralell to a and perp to c.  on parrallel surfaces to b and c i have different machined holes and grooves for various reasons and i have location callouts for the tapped holes according to what type of limits i found nessacary to accomplish the task at hand.  i have a fourth surface perp to c and angled to b so i called it datum d.  i turned my rough draft into this checker and he came back with a suggestion that i should establish the parrallel surfaces to b and c as e and f, his reasoning was if i didnt i would assume unwanted tolerance stacks and cause trouble in the asembly.  i asked him if he was talking about projected tolerance zones and he said yes.  i said that i can still do a projected tolerance zone with out making parralel datum planes.  i would like your feed back, in the application, my expereince tells me i am right.  he is asking me to go to the ansi book and show where i am right, i dont have the time, so i told him to show me where he was right.  he backed down and said he didnt know if he could specifically find any definition to what he was speaking of.  i believe he is surrenduring to my view.  i welcome any comments.
norb.
i'm trying to follow, but confused. can you please show a pic? thanks.
chris
systems analyst, i.s.
solidworks 06 4.1/pdmworks 06
autocad 06
pro,
you need to go back and correct one of the sentences that states c is perpendicular to c.  this cannot be.  you need to use spell check on the word 'parallel' too.
with three datums, you should be able to locate just about whatever features you desire on the part. including projected tolerances.
the sequence in naming the features is not important.  the sequence in the fcf is important.
a sketch would help.
making the other surfaces datums is irrelevent if they are not referenced in a feature control frame. if you are making a hole perpendicular to a datum, let's say b, but datum b can not be the surface that the feature is on, and you want the feature perpendicular to the surface that it is on, you can make that surface parallel to b and make the hole perpendicular to b and if you are concerned about fit you can include a projected tolerence zone in the perpendicular fcf.
a picture would be benificial.
forget about a b c and think of the primary datum using 3 points of a feature (no centerlines) establish your secondary (2 points) and tertiary (1) point.
all tolerance should relate to these datum’s in the order of importance. you may have positional to c b a and profile to a b c.
cheers
i appologize for the incorrect spellings.  i inherited this project from someone who has a nervous problem and by the way it was started i can see why he ahs a problem.  however this checker is not making any sense.  aardvarkdw is correct in my way of doing things.  it doesnt matter what "height" a parallel plane is to a datum it can still be related back to the datum.  i am glad to know i am not going nuts.
ring man your correct c is perp to b.
aardvarkdw is absolutely correct.  the order of the datums in the datum reference frame is the important thing, and the effects of projected tolerance zones are the same whether from the original datum structure or from the checker's proposed structure as the projection starts from the actual surface, not the datum.  adding the parallel datums means that the part must now be "re-datumed" in inspection and that is added cost without value.
proedesigner00, i'm curious why you set up a new datum-d on the angled face.  the only practical reasons that i can foresee would be (1) that whatever features are being related to that new datum-d are totally independent of the original datum structure, or (2) the setup of the new datum structure in inspection grossly overwhelms the cost of the setup and the error introduced (with respect to the original datum structure) is not critical.  if that isn't the case, the features should be related back to the original datum, and scrap the one on the angled face.
stick to your guns with the checker.
jim sykes, p.eng, gdtp-s
profile services
cad-documentation-gd&t-product development