Love old hand drafted drawings

[Bented]

One of tomorrows jobs, the drawing is dated 1988.
What is wrong with the groove dimensions?

I do realize that the dimensions are all fractional with the smallest being 1/6", I consider this a +- .062" tolerance, no tolerance is stated. The parts are liquid cooled electrical crimp fittings in 110 copper.

[Jerrythepilot]

The dimensions don't add up to an inch and a quarter? Six flats at an eighth inch are three quarters, and a sixteenth wide groove times seven grooves is seven sixteenths. Math isn't my best subject.

[PStechPaul]

It looks like the grooves and lands add up to 19/16", which is probably "close enough" to 1-1/4" or 20/16".

I'm not sure about the 1/16" saw cut in the part with a 1-1/8" bore in a 1-1/4" body. 1/16" deep would leave nothing left - is this supposed to be two pieces?

[J Tiers]

It necks down under the grooves..... look at the phantom/hidden lines.

And the saw cut is part way through.

[oxford]

It looks like the grooves and lands add up to 19/16", which is probably "close enough" to 1-1/4" or 20/16".

If it’s “close enough” where do you get or take the missing/extra 1/16” from?

[PStechPaul]

Standard tolerance is usually 0.005", so 7 grooves at (0.0625" + 0.005") plus six lands at (0.125" + 0.005") is 0.4725 + 0.78" = 1.2525". So no problem meeting all the dimensions as stated on the drawing..

[J Tiers]

1/16" is fairly standard for fractional dims, unless the title block (which we do not see) has different ones. However, that tolerance would allow (taken in isolation) the grooves to actually be 1/32" fins, and still be in-tolerance.

That is unlikely to be what is intended........

Usually, one would, in the absence of any other information, take the smallest called-out dimension to be double the tolerance. That avoids the silly technical situation listed above, where the feature could be completely left off and the part still be "in tolerance". ("I put the grooves in, but they are at plus 1/32" in depth, which is in- tolerance")

[oxford]

Standard tolerance is usually 0.005", so 7 grooves at (0.0625" + 0.005") plus six lands at (0.125" + 0.005") is 0.4725 + 1.252" = 2.0325". So no problem meeting all the dimensions as stated on the drawing..

Better check some numbers there, your math looks about as good as the drawing.

[lynnl]

What about that 1/8" Typ, doesn't that allow a little wiggle room?

[PStechPaul]

Yeah, I thought that was way off. The total is 1.2525.

[J Tiers]

What about that 1/8" Typ, doesn't that allow a little wiggle room?

It really refers to all of the features of that type being 1/8"..... Meaning that dimension is typical of all.... avoids having dimensions for each one.

The stated tolerances for fractional dims are all over the map depending on whose rules you like (unless given on the drawing, in which case that is the rule). Technically, any "rational" number is exactly stated by the "ratio" (fraction). But some are long decimals, and others are repeating decimals with no end. So there may be no actual dimension stated in a practical sense by a fraction (meaning one you can measure effectively).*

hence the half the smallest fraction... It's as good as anything else, and avoids some of the silly results.

* Yes, in the spirit of Greek geometrical construction, you can theoretically arrive at 1/3 of any distance, with dividers. But I would hardly call that "practical".

[MrWhoopee]

In most of the drawings we received, and all of the drawings I made, titleblock (default) tolerances were .xxx=+/-.005, .xx=+/-.010 and 1/x=+/-.015.

Just to play the devils advocate, the 1-1/4 dimension applies to the "course" knurl, close examination shows the right side extension line doe not align with with the edge of the last groove.

I would be ashamed to produce a drawing dimensioned like that. Even assuming +/-.015, the accumulated tolerance for that stack of fractional dimensions is ridiculous.

[Mike Burch]

What exactly is a "course knurl"? Is it just a "coarse" knurl specified by someone who can't spell, or is the hand-written legend "cross knurl" correct? Or does it just mean that cutting grooves is just a coarse form of knurling?

While the function of the grooves is not immediately obvious from the drawing (and the point of knurling the lands even less so), surprisingly, the 14 sides of the grooves do increase the surface area in that part by about a third, so if that is where it is in contact with the liquid coolant they might have some additional cooling effect.

Or is that just wishful thinking?

[PStechPaul]

I still think the saw cut goes all the way through. A 1/16" cut in a 1-1/4" diameter cylinder with a 1-1/8" bore has 1/16" walls. And the bottom drawing shows a gap. But maybe the saw cut does only go part way - that information is hidden. Yeah, now I see what might be meant.

[J Tiers]

I still think the saw cut goes all the way through. A 1/16" cut in a 1-1/4" diameter cylinder with a 1-1/8" bore has 1/16" walls. And the bottom drawing shows a gap. But maybe the saw cut does only go part way - that information is hidden. Yeah, now I see what might be meant.

Looking at the top drawing, the cut should go half-way through the tube. We see what look like solid (denoting visible edges) lines showing the thickness of the tube at the bottom of the cut, which means the cut terminates there, and you would see the bottom of the cut in that view. It would be clearer if we saw all the side view.

The drawings in past times were clear, just not clear in the same way as full-on GD&T. If the drawing is from 1988, I think GD&T was extant and in use at the time. Maybe not so much in the US, but in europe. Obviously that customer had not adopted it yet.