Hi Dan,
Very nice. Do you have a recommended tolerance on each of the dimensions? Standard three place tolerance would be +/- .010. Do your recommend tighter?
The tolerance on the 0.375 and 0.500 radius and the remaining material under the slot 0.250 could be +/- .030. Do think the total height 0.385 should be hard and not reference? I realize you are trying to avoid double dimensioning with the definition of the 1/8 NPT definition.
Thanks for making.
Cory
You are welcome.
Tolerances General: I hesitate to assign tolerances based on a small collection of used parts and one new old stock (nos) one. I would really like to have at least five (5) new old stock parts each made in different time frame to see what kind of variation the maker(s) had. This is not a functionally critical part like an intake valve so most dimensions probably didn’t mean a lot to the designer. I didn’t do an exhaustive survey yesterday nor did it get my Machinery’s Handbook out but this does not seem to be a standard “size” commercial plug, at least not currently. The designer also probably was calling out fractions and not decimal dimensions. Some issues:
1) Just in the parts I have access to there is more than a little variation is surface finish (tool mark artifacts) and crispness of cuts (hanging chip(s) and or burr(s) left behind). Surface finish and cut defects affect cut accuracy, easy of measurements (reproducibility), and repeatability between persons doing measurement. Measurement tools used makes a difference; micrometers, calipers, analog, digital, a high resolution optical comparator, or co-ordinate measuring machine (which would be tough on this little part). I don’t have access to a surface texture measurement machine anymore but from experience I would guess that non-thread machined surfaces on the parts I have range between maybe 64 and 125 rms over a 0.03 inch stroke. One used part had a roughly 0.003 inch tall rotary cutter burr left on both of the slot edges of the crown.
2) Plating details. My sketches are intended to be before plating. (It can take many hours to completely study, make conclusions, and define every detail and that assumes there is access to determine what the original plating was and how thick it is. We have been saying bright zinc for years but under a microscope this nos one looks like it might be plated in cadmium. Whatever the sacrificial coating is it is very thin. It might be in the 0.0002 to 0.0008 inch range. It does not hide tool marks.)
3) Defects. All the used parts on hand have some degree of what could be called wear and tear. That made using good lighted magnification to see exactly what I was measuring very important; an artifact of manufacture or an artifact of an installation or extraction tool. The nos part I have had a chip from the slot cutting operation jammed into the groove bottom. The chip was fully plated and there was no shadow in the groove so the chip found its way there after plating. I didn’t notice it until I got the radius gauges out and put the part under the binocular microscope do test fits. To measure slot details I had to remove the chip. If I had just been out on the bench top I don’t think any gauge would have “fit” very well.
0.385 Reference Dimension: I chose not to make it a hard number. Of all the dimensions that would be, my interpretation, the hardest for a machinist or numerically controlled machine to control. Across the parts selection I have that slotted crown surface has the most variation including most variability in surface finishes from pretty smooth out near the perimeter to pretty rough near the center. Some parts have tiny semicircle almost flats near center of the slot as an artifact of machining. That also doesn’t address the situation that it is not an even nice neat number. Most dimensions for countless parts before the finite element analysis revolution were based on being easy and typical. That was true of tolerances also. If I did a tolerance loop analysis and came up with ±0.012 I would put ±0.010 on the drawing. That did two things, 1) it was simple and 2) if gave a little wiggle room for different people measuring in different places under different ambient conditions, and with different tools. Specially if the machinist got +0.010 average and I got +0.011 average I knew the part was still okay and would in writing allow the deviation. Where the dimension worked out I can’t conclude whether the designer specified 3/8” or 25/64”. It could have been either but 3/8 would have been simple. As a non-critical dimension it really doesn’t matter I believe. I also measured a brass part that preceeded the plated steel part. The corresponding dimension came out to 0.363. I don’t have a for sure clue if the designer meant 23/64 or 3/8. It very well could have been 3/8 for both of them as a ±0.02 inch tolerance would have covered both parts. (Then you step into do you make replacement parts like the factories did or do you clean up dimensions and tolerances. My best examples are Shelby American drawings, mostly by Moore and Remington, for parts and assemblies Shelby American had to provide for new street and race Cobras. In many cases you cannot take a “factory” drawing and exactly replicate what they made / had made and actually used. Sometimes it is lack of details and sometimes the fabrication methods didn’t lend themselves to making exactly what was drawn for some reason or another.) Anyway, how many would notice a part once installed if the dimension turned out to be anywhere between 0.36 or 0.39? 0.385 just happens to be what my new old stock part measures. If I was having some made I would probably go for a 0.385 MAX just to be simple.