Common Dimensioning Symbol Errors

No matter how good a dimensioning system is — GD&T, anyone? — there will still be errors encountered on drawings, simply because there will always be human beings who are behind the creation of a new drawing.  And of course we all make mistakes.  But I want to point out a few of the more common mistakes that I encounter in my travels.

•   Failure to include a diameter symbol in a feature control frame when needed.  I’m thinking particularly of position and perpendicularity. When tagging these tolerances to a hole or pin, you usually need to include the diameter symbol before the number, so that the axis of the feature is contained in all directions.

There are times when a hole’s position tolerance should not use a diameter symbol: if you really only want the tolerance to apply in two directions.  But that must be clearly indicated by proper using of dimension arrows.


No dia symbol, so the tolerance is assumed to be two parallel planes

No dia symbol, so the tolerance is assumed to be two parallel planes, but in what direction?


Now the direction of the tolerance zones is clear, and the result is a square tolerance zone


This is probably what the original intent was: A cylindrical zone indicated by the diameter symbol



•   The next common error I’d like to review is similar to the first — using a diameter symbol when it shouldn’t be there!  I see this in feature control frames for circularity, cylindricity, circular runout, and total runout.  It might be tempting, because each of these is applied to a round feature, but the tolerance number given is NOT a distance across a circle (which is the definition of diameter), but a radial bandwidth.   This one isn’t as egregious, because there is only one way to interpret these symbols, and thus the meaning isn’t different.

•   Another error — probably the most common one I’ll be listing here — is the improper location of the datum feature symbol.  Too many times I’ll see this triangle symbol tagged to a center line, because the designer/engineer thinks that the datum will be the center line.  And that is TRUE!  The theoretical datum is very often a center line or axis.  However, the triangle symbol doesn’t identify the true, theoretical datum; it is supposed to identify the physical feature from which the theoretical datum will be derived.  Notice the difference!

While that is sometimes easily forgiven, the bigger problem is when this misuse of the datum symbol creates ambiguity, such as this one on the left, and the corrected version on the right:



•  The next error is related to quality control and statistics.  Many of you may be familiar with the practice of identifying “critical characteristics” which require the inspector to measure something and keep a log or spreadsheet for these measurements.  This allows the long-term statistical trends for that dimension/tolerance to be tracked.  This gets into things like standard deviation, six-sigma, Cpk values, and other lovely terms from statistics.

When imposing these ideas onto GD&T, however, a common mistake is to flag a basic dimension as a critical characteristic.  This shouldn’t be done, because basic dimensions themselves have no tolerance — there is nothing to track!  Instead, the real variation to be measured is shown in a feature control frame.  That’s where the critical flag should be noted.


•  Finally, sometimes you may see the position symbol used on a single feature (a hole, for instance) and then the only datum referenced is a single planar datum that happens to be perpendicular to that hole.   In other words, the only thing being controlled is perpendicularity.   Don’t use the position symbol, then!

The full explanation of this was covered in an earlier blog entry, found here.


That’s it for now — feel free to send comments/suggestions about these or other common mistakes that you encounter.  Happy spring/summer, everyone!



  1. Once datum targets have been established, is it accepted practice to dimension part features from one or more of those targets?

    • Hi Tony… I would say no, because the datum targets themselves are not datums. You can dimension part features from a datum that is created from the targets, but not from the targets themselves.
      Some people may say that this is just semantics, but if we really study the difference between a datum, datum feature, simulated datum plane, datum targets, etc., then we’ll see the difference.

  2. Talking about critical caracteristics.. can a “pass through caracteristic” be a KPC or a critical caracteristic on a drawing?

    • I don’t know of any standard that addresses “pass-through characteristic,” so there may not be any official answer.
      I’m somewhat familiar with the term: a characteristic that is not detectable to the immediate customer, and therefore a defect in this characteristic may be passed on to the next customer (OEM).
      With that in mind, I don’t see a clear-cut answer to your question. At first it doesn’t seem to make sense that a very important, critical thing would be ignored like that. But there may be something made by a Tier-2 supplier that is critical to an OEM, yet it’s not measured by the Tier-1 supplier. So it might actually be possible. I’m open to ideas on this from other folks!

  3. I see many drawings with position control frame applied to a radius that has less then 180° of an arc to it.

    Position is only for a feature of size.

    Rad with >=180° IS a feature of size,
    Rad with < 180° of arc, is NOT a feature of size.

Submit a Comment

Your email address will not be published. Required fields are marked *