Quote:
|
Originally Posted by Elgin Clock
0.050 or 50 thousandths, is larger than you can think, and may affect form fit or function a bit with robots being 3 feet wide. (If it is an overall chassis cross member dimension).
|
To put this in perspective, how many teams are sure that their frame is accurate (in two dimensions) to within 1/16" (0.0625")? (Answer: many, but not a large majority, I would imagine.) How many teams can further guarantee that their frame won't distort this much under normal use? (Answer: few.) If you're cutting with a saw, good luck--the kerf isn't always easy to judge.
Quote:
|
Originally Posted by Elgin Clock
But, to add to that, a standard dimensional tolerance is +/- .002 of an inch on the parts I make on CAD at work. (Small cable connectors about an inch long (or less) total.) Think: the connector at the end of the cable snake that invades your home and brings you your TV and cable modem services. About that size.
|
I don't think that 80-20 structural framing is in the same league. Really, 2 thousandths is insignificant for the applications that 80-20 makes sense for. I'd guess that a change of 5°C would cause about that much in thermal expansion alone.
Let's see: Δd = d
0αΔT; for a 28" piece of framing, α = 23 × 10
-6°C
-1 (we'll call it constant for small ΔT), at 20°C, in 6101 aluminum (the closest I could find to 6105, which is actually used in 80-20); Δd = (28)(23 × 10
-6)(5) = 0.0032"
So if Δd = 0.0032", we're dealing with as much thermal expansion as we have machining tolerance, merely in taking it from the unheated cargo trailer to the pit area. That's why I think 0.002" is crazy, for our purposes.