Re: [Phys-l] Sig figs

[Michael Edmiston <edmiston@bluffton.edu>]

Well, yeah, I sort of used sig-figs, but I was doing it in a pejorative 
way... to point out the the typical machinist is not going to treat 1.0 and 
1.000 much differently.  I didn't give any explicit uncertainty because it 
depends on the shop or even on the particular machinist, or what positioning 
devices are present on the particular machine used.

When I submitted drawings to the machine shop when I was in graduate school 
(long time ago), most things came back at roughly +/- 0.0005" no matter 
whether my drawing showed one decimal place or three.  (This is assuming it 
was not a large piece, say not having a dimension larger than 12 inches or 
so.)  However, one machinist routinely did +/- 0.0001 work, and another 
routinely did +/- 0.001 work.  They both worked at roughly the same speed. 
The person who routinely did the better work was 20 years older (45 versus 
25) and was simply more experienced at working quickly as well as 
accurately.

My point was, if you specify 1.0" you certainly are not going to get 1.0 +/- 
0.1 inch.  I have never met machinist who would do that.  If you specify 
1.0" you are probably going to get +/- 0.001 inch, but you might get a 
little better (perhaps +/- 0.0005) or you might get a little worse (perhaps 
+/- 0.005).   The professional machinists I have worked with would be 
embarrassed with +/- 0.005 work on small parts (several inches or less) even 
if you gave only one decimal place in your drawing.

* * * Slight aside... * * *

I mentioned in an earlier post that it is helpful to let the machinists know 
what you are up to.  If they are told that one piece fits inside another 
piece, and you would like 0.002" clearance between the two, then they will 
make it that way.  If you just give the dimensions, and they are doing +/- 
0.001 work, you could end up with no clearance.  Of course you can/should 
specify the outer piece as "minus none" and "plus something" and vice-versa 
on the inner piece.  But I always found that machinists liked talk to you 
about you are doing.  They might not understand the physics, but they 
understand the mechanical parts, and they can often make design suggestions 
you haven't thought of.  So I would visit the shop with sketches and 
half-finished drawings, and ask for advice.  Of course this only works in an 
environment where you can talk to the machinists face-to-face as opposed to 
simply submitting finished drawings.

Michael D. Edmiston, Ph.D.
Professor of Chemistry and Physics
Chair, Division of Natural and Applied Sciences
Bluffton University
1 University Drive
Bluffton, OH 45817

419.358.3270 (office)
edmiston@bluffton.edu


--------------------------------------------------
From: "Bernard Cleyet" <bernardcleyet@redshift.com>
Sent: Sunday, October 17, 2010 3:34 AM
To: "Forum for Physics Educators" <phys-l@carnot.physics.buffalo.edu>
Subject: Re: [Phys-l] Sig figs

> I think the idea is if I specify I want an inch but write 0.999 is ok and 
> he's tuning he'll stop there.  This, of course is just before the accuracy 
> available w/ position encoders.
>
> bc  doesn't remember what the UCSB machinist wanted (55 years ago)
>
> p.s.   I think you just used sig. figs.  Don't let JD know.
>
>
> On 2010, Oct 14, , at 19:56, Michael Edmiston wrote:
>
> > I don't know where your machinists came from, but here is what my 
> > experience
> > is, and I have worked with many machinists at many different places,
> >
> > If you ask for 1.011" you will get 1.011".
> >
> > If you ask for 1.0" you will get 1.000".
> >
> > Same price
>
> _______________________________________________
> Forum for Physics Educators
> Phys-l@carnot.physics.buffalo.edu
> https://carnot.physics.buffalo.edu/mailman/listinfo/phys-l