From: Henry Spencer <henry@zoo.toronto.edu>
Newsgroups: sci.space.tech,sci.space.policy
Subject: units (was Re: Space Station Construction)
Date: Tue, 19 May 1998 14:33:49 GMT

In article <35610D87.6C00DC33@elec.canterbury.ac.nz>,
Robert Lynn  <lynnrg@elec.canterbury.ac.nz> wrote:
>kilo k, mega M, giga G, tera T, peta P, exa E prefixs only!  (quick
>aside: is the prefix for tera T or t?

It's T -- there is now a consistent pattern that the big prefixes are
uppercase letters and the small ones lowercase.

By the way, you forgot zetta Z and yotta Y, and their counterparts at the
other end, zepto z and yocto y.  (If it gets hard to remember these, the
key thing to know is that starting with peta, they're twisted variants of
the familiar numeric prefixes, e.g., peta comes from penta.  They're
twisted a bit to avoid confusion and give them unique initial letters.)

They sound a bit silly, but they do have uses, in principle -- the local
intergalactic distances are conveniently expressed in Zm and the size of
the universe in Ym.  Of course, it may be a little while before the
astronomers can be convinced to switch from parsecs and light-years...

>N/mm^2 is of course equivalent to
>MPa which every engineer is familiar with, so why not just say MPa?  It
>is quicker and easier and reveals that the author is at least slightly
>aware of technical conventions.

Remember, the paper I was summarizing is 20 years old.  I didn't convert
partly because I was feeling lazy and partly because it's usually better
to avoid gratuitous unit conversions when quoting somebody else -- there
is less possibility of error if conversions (even just of unit names) are
delayed until calculation is actually necessary.

>[Mod note: in case anyone seriously does not know what KSI are,
>it's "Thousands of pounds force per square inch".  100 KSI yield strength
>is 100,000 lbf/in^2, or 689 MPa.  There is no group policy on the
>use of SI or english units, but the moderator is stubborn.  -gwh]

Next he'll be using stones/furlong-fortnight^2. :-)
--
Being the last man on the Moon                  |     Henry Spencer
is a very dubious honor. -- Gene Cernan         | henry@zoo.toronto.edu

From: henry@spsystems.net (Henry Spencer)
Newsgroups: sci.space.tech,sci.space.policy
Subject: Re: units (was Re: Space Station Construction)
Date: Tue, 2 Jun 1998 13:07:31 GMT

In article <6kvmld$n2i@lace.colorado.edu>,
Frank Crary <fcrary@rintintin.Colorado.EDU> wrote:
>>I agree.  I find yottagrams to be a natural unit for remembering
>>planetary masses...
>
>Personally, I don't like it, but I may have some biases because I'm
>a theorist. For many purposes (e.g. orbital dynamics) it is easier
>if you use nondimensional units...

While it makes things easier at theory time, it can make life harder later
on.  Bob Forward, in his JBIS paper (Nov 1989) on spacewarps, comments:

  Determining the correct equations for the physical magnitudes of the
  quantities being discussed was found to be difficult.  All the papers
  by mathematicians about [general relativity] were written in units
  where h=c=G=1, while papers by nuclear physicists about models of the
  electron and nucleus were written in mixed cgs units where the constants
  for the permeability and capacitivity of space are suppressed.  The
  proper procedure for the reinsertion of the various constants to form
  consistent equations for the numerical calculation of the magnitude of
  any expected experimental result is not obvious, and there exist two
  papers (deliberately not referenced) with drastically different
  versions of the same conversion...

>added advantage, you don't get any problems or confusion from planets
>having masses that are 10 to the twenty-somethingth power. (E.g.
>Jupiter is roughly 300 Earth masses; Mars is roughly 0.1; or Jupiter
>is about 0.001 solar masses.) So you get units of mass that are convenient
>(i.e. within a factor of 1000 of one) ...

There is also a more subtle issue here, which Peter didn't raise when
talking about yottagrams:  planetary masses are another one of those
troublesome areas where we've got a poorly-known calibration constant
standing between the measurable values and the standard units.  We measure
planetary masses using gravitational effects.  The effects can be measured
to six or seven decimal places without great difficulty... but to convert
them to masses we have to use G, the constant of gravitation, and G is
known to only about four places.  (Gravity is such a weak force that G is
quite difficult to measure precisely.)

High-precision work, e.g. for interplanetary navigation, uses values of
GM (aka mu) -- the measured quantity, and also the one of interest for
calculating gravitational effects -- usually quoted in km^3/s^2.  The use
of km instead of m, and the rather small value of G, give the numbers
halfway convenient sizes, e.g. the GM of Mars is about 42828 km^2/s^2.
--
Being the last man on the Moon is a |  Henry Spencer   henry@spsystems.net
very dubious honor. -- Gene Cernan  |      (aka henry@zoo.toronto.edu)

From: henry@spsystems.net (Henry Spencer)
Newsgroups: sci.space.tech,sci.space.policy
Subject: Re: units (was Re: Space Station Construction)
Date: Fri, 5 Jun 1998 14:27:10 GMT

In article <pgISDTAk$wd1EwjD@branta.demon.co.uk>,
Del Cotter  <del@branta.demon.co.uk> wrote:
>>Why yotta_grams? Use tons (tonnes? I know English-speakers also have
>>'short' tons and 'long' tons, and the spelling is mysteriously connected
>>with that, but since we are talking metric...).
>
>It is not.  If you're talking metric, you *must* say tonnes, since a
>tonne is 1000kg and a ton (of any variety) isn't.

There is long precedent for calling 1000kg a metric ton or just a ton;
even my 20-year-old dictionary lists both of these usages.
--
Being the last man on the Moon is a |  Henry Spencer   henry@spsystems.net
very dubious honor. -- Gene Cernan  |      (aka henry@zoo.toronto.edu)