Re: Round-off error (Long) Message #13 Posted by bill platt on 24 Aug 2003, 7:01 p.m., in response to message #7 by bill platt
Hi Torsten, r.d., Nelson et al:
Well, I did wink when I said, "All good designs should be done in INCHES." Though it still is an interesting irony that it is "easier" to work in inches if you know you will need two unit systems in the end.
Also on this topic, I worked for a EU shipbuilder for some time, and as it was a new start-up yard in the US, the Euros set us up to take a number of classes as training: AutoCAD, various ship design programs, German language (this was actually a great benefit--it made the business trip to Deutchland much more interesting, fun and meaningful), and: "metric"! The teachers were all americans (yes, one was Canadian) and really quite good; but what we learned in "metric" class (we all rolled our eyes when we heard we had to take this one!), what we learned was that contrary to popular belief, Americans ARE bilingual---that is to say, inches, and Metric! (We all used SI units in school, grew up in the 70's when the Carter administration had put wheels in motion to switch over, and then the first things I remember in the Reagan epoch were: 1: can metric, 2: bring back the (already obsolete) B-1 bomber program, 3: eliminate federal funding for Solar Energy (and probably wind etc), 4: remove sound decibel rating/warning stickers from lawnmowers). Never mind all the other good things that did come along.....(too off topic!)
So, yes, metric is, well, too EASY! like, you never have to play games with the factor 12 or .083333333333333333333333333333333......
Another couple good tidbits: the first is that in conversion, there are two concepts: "hard" conversion, and "soft" conversion. I am sure I will mix the terminology up as I remember it being counter-intuitive, but the idea is this: an "exact" conversion is "soft", e.g. a 4" flat bar converts to a 101.6 mm flat, whereas a "hard" conversion rounds (to some predetermined pseudoprecision) to 100 mm. This sort of stuff is important for concept and preliminary design, and for picking "equivalent" parts etc.
A lucky coincidence:
In ship design ("naval architecture"), we design in "long tons" not in little mass units like slugs or lb-m or kg. A "long ton" is 2240 lb, (a "short" ton--the landlubber's ton--is 2000 lbs). By wonderful happenstance, a metric ton (or tonne to make it sound all frenchie--the frenchiness of international standards is a whole 'nother topic, like in radio Morse code, N3LPX DE F1FRE FBOM, 73 F1FRE SK) is 2204.623.... lbs. And so it is both confusing (metathasizing the "4" from the 1's column to the 10's column) yet also simplifying---because 2240/2204 = only 1.6% difference, and so for general design concept comparisons stuff, they (metric and long ton) are the same.
The last tidbit is that all U.S. Federal highways are engineered in meters----and have been ever since the Carter administration. Evidently, they (the DOT) made the switch early on, and when Reagan said, "we quit" the DOT said "not us--we're sticking with it!"
Regards,
Bill
Edited: 24 Aug 2003, 7:05 p.m.
|