Scientific paper on accurate (proper) measurement of tight tolerance bores? - Page 2
Close
Login to Your Account
Page 2 of 2 FirstFirst 12
Results 21 to 23 of 23
  1. #21
    Join Date
    Feb 2014
    Country
    CANADA
    Posts
    37
    Post Thanks / Like
    Likes (Given)
    4
    Likes (Received)
    17

    Default

    Ok some nerdery here. In no way is this meant to be much more than a thought dump - you're an extremely competent guy Seth and I'm sure much or all of this has occurred to you already.

    First, a practical solution is to request that the customer either provide you with approved gauge pins or approve ones you send them. That way there can be no debate that the metrology means itself is the problem. If they're rejecting already made parts as out of spec based on this, that's a tough one, and my suggestion would be to chat to the design engineer and ensure that their stated tols have a specific purpose. If an 0.1251 gauge pin doesn't fit in there, nor might they have sufficient clearance for the 0.1251 upper limit on the mating part, etc. It might also be the case that they're able to revise their tolerance rather than reject the part.

    The full analytical solution requires (much) more definition, because there basically is no limit on how far down you can go with measurement resolution, and in many regards the measurement resolution needs to greatly exceed the part tolerances by several orders of magnitude. I am yet to see a drawing of anything (ever) that is actually fully constrained in all degrees of geometry for all features - it just isn't possible, nor is it practical to measure any geometry that fully (the number of data points required for starters is totally implausible). This goes for both the part and the metrology equipment.

    Let's consider an ideal metrology scenario - your Go gauge pins are classically perfect (meaning all measurements consistent to within the width of the constituent molecules of the material). "Classically perfect" is already limited in its definition, because even though it's already at a totally unrealistic standard of measurement for machining parts, it's also ignoring any non-Newtonian physical effects like probabilistic variations of geometry (technically every particle position is just the average predicted by probability distributions, not a hard and fast "fact"... so it's constantly varying). Anyway, our classifcally perfect pin has a consistent diameter at nominal dimension, is absolutely axisymmetric, absolutely consistent radius (no lobe profile), zero taper, zero eccentricity draft, Ra = 0.5Rz = molecular width etc, Young's modulus of infinity (absolute rigidity, no effect of measurement calibration tools on measured diameter of gauge pin), thermal expansion coefficient of zero (or temperature gradient of zero and temperature delta of zero between tool/part/atmosphere).

    In this case, if you had the equivalently perfect characteristics in your part feature OTHER than having a real-world Youngs modulus (no infinite rigidity) and the diameter being exactly at the lower limit of the tolerance, you would be able to fit the Go gauge pin in. The reason you'd be able to fit it in is that the deformation of the part material would allow it, because there'd be no geometric (mechanical) interference and any static pressure (even the air trapped in the hole) would be sufficient to create enough elastic deformation for there to be clearance. However, you'd see significant resistance inserting the pin from the viscous boundary layer effects of the air lubricating the hole, and getting it started would require a taper on the pin (and be a nightmare anyway). As soon as anything else has a non-zero tolerance (cylindricity, perpendicularity, surface profile, circularity, straightness etc - and yes some of these are compounds of others) all that goes out the window because you end up with a massive matrix of codependent confounding variables, and generally speaking it's utterly impractical to measure any of them sufficiently well that you can actually calculate what that means for your limits. As a result, you need to be able to measure many more geometric aspects to somewhere approaching atomic levels of accuracy in order to really be able to analytically determine what size gauge pin fits in your hole.

    This is of course basically just theoretical nonsense that does not help with what any of us are dealing with in reality, but I find it interesting to consider the extreme limits of precision in that regard. Metrology nerds will have better established empirical methods of establishing practical approximations of accurate measurements, but I would posit that there is no truly "proper and accurate" measurement system for such features if we're going to be pedantic about it.

    However, agreeing with the customer on which gauge is acceptable as a Go and what is acceptable as a No-Go (not just a nominal size, but a physical unit they have signed off on) or holding a tighter tolerance to ensure you're well inside the brackets is eminently more practical in my opinion.

  2. #22
    Join Date
    Feb 2021
    Country
    IRELAND
    Posts
    1
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    0

    Default

    I feel your customer is technically wrong but right in spirit and in practice as well.

    Deltronic class X gage pins (assuming this is what you are using) have a rated tolerance of plus 0.00004", minus 0"
    Form error can be ignored as both pin and hole should have perfect form at MMC (I assume ASME standard given the imperial units)

    If the hole accepts a .12500" pin but not a .1251" one, all you can infer is that the hole size is:
    greater then .12500"-0.00000"
    smaller than .12510"+0.00004"

    So all you know is that .12500 < Hole < .12514

    This mean the hole might be within specification or not, (and is quite more likely to be out)

  3. #23
    Join Date
    Jun 2012
    Country
    UNITED STATES
    State/Province
    Michigan
    Posts
    13,142
    Post Thanks / Like
    Likes (Given)
    4459
    Likes (Received)
    4741

    Default

    Here find a 2018 catalog:
    https://www.deltronic.com/literature...Guide-2018.pdf

    Back in the day that I made one-up/fewup gages, I could easily spend an hour on a set.
    Often I would also make one with only two contact points to check out of round.

    Mostly used a 13 grinder (B&S 13).


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •