MetalworkingFun Forum
Dimensioning Question - Printable Version

+- MetalworkingFun Forum (http://www.metalworkingfun.com)
+-- Forum: Machining (http://www.metalworkingfun.com/forum-5.html)
+--- Forum: General Metalworking Discussion (http://www.metalworkingfun.com/forum-6.html)
+--- Thread: Dimensioning Question (/thread-2399.html)

Pages: 1 2


RE: Dimensioning Question - DaveH - 12-10-2014

Tolerances have two main purpose's
One is so the bits and pieces fit together where ever they are made and secondary the machinist making the part doesn't have to "nail it" - it is time consuming and hence expensive.
In the home workshop tolerances are not always critical because we can make the parts fit and time isn't the prime concern, except of course when it's tea time Big Grin

Smiley-eatdrink004 
DaveH


RE: Dimensioning Question - TomG - 12-10-2014

(12-10-2014, 08:16 AM)Mayhem Wrote: Tolerance is still something that I am not fully comfortable with.  I must admit that when I watched Tom's latest videos it seemed very strange to hear him say a part was -0.000 and +0.003 and so he was aiming for something in the middle.  Even stranger to hear him say that if the dimension was only to two decimal places he wouldn't even bother to take a final measurement.

Of course, having him explain why made a lot of sense but it just seems strange when I have been trying really hard to nail a dimension exactly, when in hindsight, I probably didn't need to worry.

That's really good to hear Darren. Part of good design is to make use of tolerancing in a way that makes the part easier and less expensive to make, but still meets it's intended purpose. That's why some of the dimensions in the knurl holder, such as the length are two place, and the slots, outside dimensions and hole which mate with other parts are three place, or three place with an attached tolerance. For instance, if a 1" thick part has a general tolerance of plus or minus .005" and you spend extra time machining it to within .001", you have wasted time. If the part was properly designed, it will work just as well at 1.005" as it will at 1.000". Capisce?

A video on the application of tolerancing is one of the topics on my list.

Tom


RE: Dimensioning Question - rleete - 12-10-2014

(12-10-2014, 08:29 AM)DaveH Wrote: Tolerances have two main purpose's
One is so the bits and pieces fit together where ever they are made and secondary the machinist making the part doesn't have to "nail it" - it is time consuming and hence expensive.

I (and I suspect many hobby machinists) don't work too much about tolerances.  I make a cylinder to roughly the right bore, and then make the piston to fit.  If it's .50 or .55 doesn't really matter.  It's the fit that's important. 

In production shops it's different.  You may not even know what the mating part is.  So, tolerances make it fit when you don't have the time to custom fit each part, and it's also what makes interchangable parts possible.  Time is money in a shop.

Tolerances, knowing how tight to make them, and knowing where and when to use them is the mark of an experienced designer.  I wouldn't have known how to use them properly 20 years ago.  Now, it's second nature.

BTW, I work at a place that does diamond machining for optical tooling.  Those guys work to insane tolerances.  Minutes of angle and tenths of an inch are pretty common around here.  I have actually seen specifications that list tolerances in microns.  That's 40 millionths of an inch.


RE: Dimensioning Question - DaveH - 12-10-2014

(12-10-2014, 09:39 AM)rleete Wrote: BTW, I work at a place that does diamond machining for optical tooling.  Those guys work to insane tolerances.  Minutes of angle and tenths of an inch are pretty common around here.  I have actually seen specifications that list tolerances in microns.  That's 40 millionths of an inch.

Mmm ........ 'and tenths of an inch' seems a bit big or am I missing something?

Smiley-eatdrink004
DaveH


RE: Dimensioning Question - JScott - 12-10-2014

I think rleete is referring to "tenths" as in 0.0001" or one ten thousandth of an inch.


RE: Dimensioning Question - rleete - 12-10-2014

Correct. Since nobody considers +/-.005 very close, we really don't specify anything larger than 1 or maybe 2 thousandths. "Tenths" is ten tousandths of an inch.

Stupid, but oft repeated, joke around here is that "2 tenths is for camping".


RE: Dimensioning Question - TomG - 12-10-2014

Oh yeah? I can beat that. Big Grin

We make a molded polymer part that has 96 tubes in it. The bottom of each tube has a .100" x .130" piece of polymer film attached to it and each one of those pieces of film has a grid of 96 holes drilled in it. The holes are each 8 micrometers in diameter and the tolerance on the diameter is ±1 micrometer. In Imperial terms that's ±38 millionths of an inch. Now you all know why I'm so screwy. Blink

Tom


RE: Dimensioning Question - EdK - 12-10-2014

(12-10-2014, 05:37 PM)TomG Wrote: .... Now you all know why I'm so screwy. Blink

Tom

That sure does explain a lot Tom. Chin

Big Grin

Ed