Slashdot Mirror
NIST Wants An Electronic Kilogram
Dearing writes "According to the Global Engineering Journal, NIST, those not-so-standard standards people, want to give up the hunk of metal they've been calling a kilogram, even though it never weighs the same twice. In it's place, an electronic kilogram could act as the permanent standard."
NetInfo connection failed for server 127.0.0.1/local
(one of) NIST's own web page(s) on this is at http://www.eeel.nist.gov/811/elec-kilo.html. There's a lot more technical detail there than at the link given in the article.
This really does make sense to replace the artifact with something independent -- they have a bunch of "voodoo" every time they measure the current kilo to try to get the same answer.
that wouldn't work - after all, then you've just scaled the problem down to ask "whats the weight of a proton"
remember that the base physical units have to be directly related not to theor, but to empirical observation. That's the difference between "units" and "physical quantities"
MASS is a physical quantity. "kilogram" is a "unit" of that quantity. defining it in terms of the "mass of a proton" makes no sense because thats essentially a *circular* argument.
if you;re gonna construct a vast edifice of science, the foundation better be damn rigorous! this isnt just semantics, its essential, the way that we have to be absolutely sure that 2 + 2 = 4 (which can be derived from the Completeness property of the Real number Set). A good reference for basic units and quantities is here.
Don't blame me - I voted for Howard Dean. http://dean2004.blogspot.com
According to NIST [nist.gov], they've got a variance on the order of 3% per century in the observed mass (probably measured by weight) of the
standard kilo brick.
3 percent??? Do you have any idea how HUGE a variation that would be in a primary standard? Maybe if they polished it with a belt sander before every measurement...
The link you provided says:
[...]are causing the mass of the kilogram to vary by about 3 parts in 108 per century relative to sister prototypes.
Now I'm not certain about this, but I'd wager that the "108" is actually a "10 to the 8th power" that got mangled somewhere in the conversion to HTML. If so, it would represent a more plausible 0.000003% per century variation.
Second: The second is the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the cesium 133 atom.
As mass and energy are equivalent, can't they just define the mass as a certain number of barrels of that cesium 133 light? Right now they just throw away that light each second, but they could recycle it.
why, exactly, a kilogram [...] never weighs the same twice? MouseR
Presumably, because of engineering imperfections in the (mechanical) measurement devices, and perhaps also due to local variations in gravity, caused by tectonic forces, tides, etc.
The bottom line is: weighing a physical chunk of metal is as poor a standard as measuring the length of a chunk of metal. We do better if we can relate these standards to invariant values derived from basic physics.
-- We all have enough strength to endure the misfortunes of other people. La Rochefoucauld
The ultimate goal is to derive all measures from
the fundamental constants of physics.
The two most popular are "c" the special of light
and "h" Planck's quantum of action.
A recent Physics today suggests a using
E=mc^2 and E=hv, where v is a frequency.
Frequencies are the most accurately measurable
item in the universe, at a current accuracy of
one part in 10^19. So the proposal is to choose
a "kilogram frequency" that precisely defines
the kilogram. There is already a "meter frequency"
that precisely defines the meter length in terms
of light velocity. And a "second frequency"
which some frequency count close to an astronomical
second.
The least well-known constant is the gravitational
constant, measured only to four decimal places.
The probably is instrumental error, because
everything pulls on everything else.
At least twice in the past decade someone has
proposed changing the law of gravitation because
of funny measurements, but every time an
experiment error was found. The constant "G"
doesn't fit into many physics equations,
so it isn't as easy to bootstrap equations
as for the other constants and measurement units.
"NIST, those not-so-standard standards people, want to give up the hunk of metal they've been calling a kilogram, even though it never weighs the same twice."
of course it weighs different every time, it's a standard kilogram, which is a measure of mass. the weight of the Kg will differ as gravity differs - which is a fun little trick having to do with the mass of the earth and the nearby celestial bodies.
the whole point of the new measuring device is (basically) to more accurately measure the force of gravity on the standard mass - by doing some magic with a magnet keeping the whole thing in balance. this is really just getting at a better measurement of gravity than anything else.
the crux of the situation is that the only standard for a kilogram is the actual lump of platinum itself. other things, like the standard second, are based on fun stuff like exactly how many times a cesium atom vibrates at a particular temperature. it might be fun to try and define a kilogram as Exactly This Many platinum atoms and be done with it, but that's kinda tricky for the moment.
it might be a better "standard" to accelerate the "standard" mass at a "standard" rate and measure the forces. say, by swinging the thing around in a calibrated centrifuge at whatever we're calling one Gee. then you can get to the bottom of the whole "weight" issue (in terms of newtons, i suppose).
besides, unless the standard mass is made of something that's decaying (radioactively - it's not like they'd make the thing out of, say, beef), it'll be pretty much the same mass for quite some time. it's just those nitpickety scientists at the NIST (on which i read a very interesting article recently, i believe in National Geographic Magazine) who want it to be defined in terms of something that will never change
and secondly, since when is the NIST "not-so-standard"? they are the national frickin' institution for the damned things, so they should be an authority on the subject...
- Entertaining Bits from the Ancient Kernel Tree