Limits and Fits
Limits and Fits
(OP)
I'm dimensioning a bunch of 2D drawings. Some of the dimensions require tolerances where parts fit together (we are using H7/g6 locational clearance). This part I understand for the most part from reading Machinery's Handbook. But another designer added a symmetrical tolerance to the hole centre location of +-0.005 mm. How did he get this value? Is there any other good resource for learning about tolerances?
Follow @mrmitche_ on Twitter!





RE: Limits and Fits
You still have to locate the parts based upon their intended function. The tolerance will be up to the designer to dictate. Do you have control over the manufacturing process? If so, I've seen some design documents specify that holes are to be "match drilled at assembly" or variations of that phrase. This basically gives you a much freer locational tolerance but applies a process that hopefully guarantees a much tighter in-line condition, as would be appropriate for a locating feature such as a dowel pin and bushing.
However, if you're using a slotted bushing and round bushing with your pins, maybe you can dimension and tolerance them separately with +/- 0.2mm and things will simply mate up satisfactorily depending on the desired assembled condition. Or maybe things are tight and the geometry is not forgiving, and +/-0.005mm is all you can get away with.
I don't know what your features are - I speak of bushings only for illustrative purposes to convey how you would have to approach things differently.
_________________________________________
NX8.0, Solidworks 2014, AutoCAD, Enovia V5
RE: Limits and Fits
Follow @mrmitche_ on Twitter!
RE: Limits and Fits
I understand that there is a symmetry specification in ASMY Y14.5, but I have no idea of why I would use it. Maybe your designer is being silly. Symmetric about what?
If your H7/g6 fit is being used to locate something, almost certainly, you need to apply positional tolerances somewhere. A positional tolerance plus symmetry may be a way of showing that one datum is more critical than the other one.
--
JHG
RE: Limits and Fits
_________________________________________
NX8.0, Solidworks 2014, AutoCAD, Enovia V5
RE: Limits and Fits
If you want to be doing it even remotely properly you normally need to be determining what is the loosest tolerance that will ensure part function (most typically that the mating parts will fit together)*. For instance ASME Y14.5M-1994 gives some equations for making sure hole patterns match up in appendix B.
Secondly you need to ensure that the required tolerance you've calculated are within the process capability of how the part can be made and can be achieved at acceptable cost.
Just blindly copying other designs, or over relying on the tolerance block or on some industry standard of process capability is not doing it properly.
* Often the tolerance necessary for function/fit are very loose, in this case it's not necessarily worth doing in depth tolerance analysis - just pick a number well within process capability and carry on with your day.
What is Engineering anyway: FAQ1088-1484: In layman terms, what is "engineering"?
RE: Limits and Fits
Let's actually answer your question.
You need a copy of the standards. I strongly recommend getting a copy of ASME Y14.5-2009 Dimensioning and Tolerancing. GD&T is a language, not a procedure. You need to understand what you are telling people.
Most if not all of what KENAT is saying, above, is common sense. Your part has to work. I don't see the specific combination of h7/g6 anywhere in my Machinery's Handbook, but it is valid, and it sounds like an accurate sliding fit. If that is what you need, you are doing it right.
Symmetry specifies geometry in 2D, not 3D. If you are centreing something, you need either true position, or concentricity.
--
JHG
RE: Limits and Fits
Follow @mrmitche_ on Twitter!
RE: Limits and Fits
So Plate 2 doesn't require a more accurate tolerance on the hole location because it has 0.5mm clearance around the insert. But Plate 1 is guiding the shaft with the sliding fit, so the hole location needs additional accuracy?
Follow @mrmitche_ on Twitter!
RE: Limits and Fits
_________________________________________
NX8.0, Solidworks 2014, AutoCAD, Enovia V5
RE: Limits and Fits
20.00±.005 controls the position of your hole. If this is millimeters, this is very accurate, and possibly not fabricatable.
The specification Ø18 H7 is a valid thing to do on a drawing, as per ASME Y14.5. You can work out the actual tolerance by reading the Machinery's Handbook. Your machinist can work out the tolerance by reading the Machinery's Handbook. Your inspector can work out the tolerance by reading the Machinery's Handbook. I like to call up fit classes on my assembly drawings, where it functions as design information. I want numbers on my drawings. Does your CAD's tolerancing feature understand this? I know SolidWorks does.
--
JHG
RE: Limits and Fits
Follow @mrmitche_ on Twitter!
RE: Limits and Fits
There is a process for working out the size tolerances and the positional tolerances required to make clearance holes clear bolts and screws. I have written up on this, and posted it to my personal website. None of this applies to H7/g6 holes, as they provide insufficient clearance to clear multiple features. Common practice with dowel pins is to provide one round hole, and one slot.
From my Machinery's Handbook, I see that your Ø18 H7 hole is in fact Ø18.018/18. This is what you should show on your drawing. The H7 bit is design information that the fabricator and inspector do not need to know.
The 20mm dimension appears to locate your hole. Since your fit is very accurate, this locates the mating piece in your assembly. Read KENAT's remarks above. If you want your parts accurately located, you are doing it right. I am not sure your fabricator can locate a hole to ±.005mm.
--
JHG
RE: Limits and Fits
My mind visualizes your descriptions of two parts located by pins with a sliding fit. Unless the 20mm dimension from surface (side) to the hole is related functionally, so that this +/-.005mm tolerance is required (at best it would be quite expensive if possible to fab as mentioned above); or is it that the hole should be defined as a datum feature used in a DRF which would be used to locate the side surface relationship? If so then the +/- could be more liberal.
Sure would be nice to see both part dimensions. Minds do wander
JNieman,
True bilaterial is +/- tolerance however both values do not need to be the same (e.g. -0.1 / +0.3).
Maybe Solidworks uses "symmetric" as a crummy way of indicating an equal bilateral tolerance?