Symmetric dimensioning

[piznal]

A simple situation, but I have not found explicit reference to it:
- Entirely symmetric part with many mirrored features.
- I believe functional (mating) emphasis requires horizontal dimensioning relative to a part center-line (planar primary and secondary datums, with tertiary datum plane of symmetry passing through the axis of a central datum FOS).

Is a basic dimension from center-line to feature required for each feature AND its mirorred counterpart, or is it acceptable to dimension in one direction from center-line with a "2X FEATURE and feature tolerance control frame" callout, with the dimension to counterpart on the opposite side of center-line being dictated by symmetry?

Any help or comment is appreciated.

 

[drawoh]

piznal,

On symmetric parts, I usually show the dimension across the symmetrical faces. These are real, measurable dimensions, unlike dimensions from centrelines.

--
JHG

 

[KENAT]

I'm assuming ASME Y14.5-1994

If you've set up your datums properly, and are using GD&T/FCF correctly then I'd say the 2X approach is just fine. (Remember center planes can't be datum features as such, the datums need to derived from FOS - which I think you already said you were doing.)

The alternative (for features located by GD&T) is instead of dimensioning to the center line, dimension symmetrically across the center line from feature to feature. This is shown in figure 5-4.

(Some folks argue this isn't explicitly clear and you're assuming symmetric and that this figure is a case of an incomplete one, however it looks good to me.)

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[piznal]

I am definitely using a central through hole FOS (together with the planar datums) to locate the plane of symmetry (horizontal origin).

My impression of using dimensions that span between the symmetric features is that doing so implies that in the inspection process you are physically measuring from one feature to the opposite. The problem then is that the location of each feature has (presumably) been given a cylindrical tolerance zone, but that which of the two features is to be considered the "reference" for placement of the tolerance zone of the other is not stated, nor is how the "reference" feature is placed/toleranced (presumably relative to the center-line). However, I am no expert in GD&T.

Thanks for the posts.

 

[KENAT]

The dimensioning from feature to feature across the center line only works if you're using basic dims & appropriate GD&T (such as position or surface profile). Certainly doesn't have to have a circular tolerance zone.

The FCF invokes the relevant datum structure, so no your aren't directly dimensioning from one feature to the other.

You're giving the theoretically perfect location with the basic dims, then the FCF invokes the datum structure and the permissible variation from that perfect location.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[piznal]

Thanks for the feedback...so that I get this somewhat correctly, is it true then, that if a basic dim.,D, is given feature-to-feature across the center-line, there is an implied basic dimension from center-line to each feature that is 0.5D, subject to whatever positional tolerance I specify? Otherwise, if some positional tolerance applies to each feature of the pair, what is to be taken as the perfect location about which the tolerance zone is placed. Or, in other words, each feature can land within a specified tolerance zone about a perfect location given by a basic dimension, but the only basic dimension given is from one feature (or center of tolerance zone) to another, where neither feature (or center of tolerance zone) is explicitly placed relative to anything else (i.e. the center-line).

Sorry. I'm a bit obsessive, and this kind of thing has been tripping me up.

Thanks again.

 

[KENAT]

Take a look at the figure I mentioned.

My take is that no you don't explicitly need the 'centering' dimension if you are showing clearly symmetrical across the datum and the datum is referenced in the FCF.

A few people argue otherwise.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[piznal]

Got it. Thanks much.