【转帖】straightness vs flatness

yang686526

straightness vs flatness
we are currently having a discussion on straightness vs. flatness. here is the scenario: we have flat mounting surfaces that we need to check in respect to 鈥渇latness鈥? we currently use a 鈥淯鈥?shaped block that the part surface can slide into. the gap is set to material thickness at mmc and the tolerance for 鈥渇latness鈥?added.  is this best called out with straightness or would flatness be better? my thinking is straightness because of the mmc call out on the material thickness. any ideas?
find a job or post a job opening
i've always associated straightness with wires and shafts.  however, i do not have any hard documentation to back up this thought.
"the gap is set to material thickness at mmc and the tolerance for 鈥渇latness鈥?added."  flatness is not additive to the limits of size, it's a refinement.  if your "u" shaped block is intended as a flatness check gage, then it's not set up to check the flatness requirement.  a part that came in at lmc on the thickness would be allowed more flatness variation than a part produced at the mmc size because your check boundary is fixed at mmc + the flatness tolerance.
regarding the question flatness vs. surface straightness, flatness is a 3d form control of the surface.  surface straightness is a 2d form control of the surface only in the view in which the straightness control is applied.
for most 3d surfaces, the design requirement is usually that the entire surface be controlled, so flatness would be the tool of choice in controlling the variation.  
my two cents
gdt_guy
i might be mistaken but it sounds like you are checking parallelism not flatness.
we typically check flatness with an indicator on a height gage. the part is supported on the surface in question (this is a must) and we zero the indicator at one point on the surface. the entire surface is scanned and the fim is recorded as the flatness. this is not an easy task.
because flatness is not (cannot be) applied to a feature of size i don't think you can correctly use mmc or lmc. because of this i do not think flatness refines the size limitation at all.
straightness makes more sense to me but i would argue that parallelism might be more appropriate.
jec3300,
what you describe doing with the u-shaped fixture is no-go gage for straigthness of a feature of size (fos) at mmc.
the flatness of a fos surface cannot exceed its size tolerance. rule #1 (perfect form at mmc) prevents this from happening. for instance, if the a bar is .25卤.01 thick, the flatness callout cannot be more than .02 (.01+.01). as gdtguy said, flatness is a refiment. in this case you no-go gage width will be .26 (.25+.01).
straightness can be specified to a fos. in this case, rule #1 does not apply. it is one of the exceptions to rule #1. if the straigthness call out is specified at mmc, a fix gage like your u-shaped fixture can be used to inspect the part. for instance, a bar .25卤.01 thick can have a straightness at mmc call out of .05. the width of your u-shaped fixture will be .31 (.25+.01+.05). the feature control frame must be placed inmediately below the thickness dimension. straightness is a 2d call out and it only applies in the direction of the drawing view showing the feature control frame . if you are interested in straightness in both directions (like flatness), another fos straightness requirement must be specified in the other direction.
i think i was/am incorrect in a previous post. the flatness specification does refine the size limitation as others indicated. i should have thought more about my response before posting, my (big fat) mistake.
sorry if i caused any confusion!
just the fact that you are using a gauge such as this indicates that you are checking a feature of size. flatness is never applied to a feature of size. if i understand what you are describing, you are most likely checking straightness whether you know it or not. in this case the gap should be the mmc of the feature of size and no more. you said something about flatness tolerance added. re  
straightness may be applied to 鈥渟lab like鈥?feature of size surfaces on either a rfs or mmc basis. paragraph 6.4.1.1.3 of asme y14.5m -1994 explains the condition and states that the 鈥渄erived median plane鈥?of the opposing surfaces must remain between two parallel planes spaced the tolerance value apart. naturally if the tolerance is specified with a mmc modifier the tolerance is variable with respect to size and there is either a max or min virtual condition boundary depending on whether the features of size surfaces resemble a cavity or a slab. as jv1 stated the specification for straightness must be shown beneath the dimension rather than attached directly or with a leader line to the relevant surface(s). when the callout is applied to the feature of size dimension however, it is not directionally dependent as it constrains the derived axis or derived median plane within a cylindrical or parallel planes zone, therefore it is not necessary to show the orientation of the constraint as it is with other straightness callouts.
an example of the use of straightness at mmc for a flat 鈥渟lab-like鈥?feature of size can be found by examining recent product designs for automotive clutch pack plates. the plates, either friction or separating, historically have a specification for size and notes constraining the variation in size and the individual effective overall thickness (size and form) stating:
鈥淢aterial thickness variation x.xx max per piece鈥?br />鈥淢ust pass freely between parallel plates spaced x.xx apart鈥?
recent designs have replaced the latter note with a straightness at mmc callout depicted beneath the size tolerance (where the mmc size plus the value for the straightness specification equal the former tolerance for passing between parallel plates) but no one has yet found an equivalent way to specify the material thickness variation with the asme standard without a note.
you will find many designs that have translated these historical notes into a combination of size, flatness, and parallelism with some retaining the 鈥減ass between鈥?note as well. commonly you will see that the parallelism callout has been specified with a smaller than the flatness callout on the opposing surface in an attempt to address material thickness variation (which it does not do).
paul
paul,
thank you for correcting my mistake. as paul correctly stated "when the callout is applied to the feature of size dimension however, it is not directionally dependent as it constrains the derived axis or derived median plane within a cylindrical or parallel planes zone, therefore it is not necessary to show the orientation of the constraint as it is with other straightness callouts." therefore, only one feature control frame is needed.