Slashdot Mirror
The Future of the Kilo: a Weighty Matter (theguardian.com)
A lump of metal in a building near Paris has long served as the global standard for the kilogram. That's about to change. From a report: Later this month, at the international General Conference on Weights and Measures, to be held in France, delegates are expected to vote to get rid of this single physical specimen and instead plump to use a fundamental measurement -- to be defined in terms of an electric current -- in order to define the mass of an object. The king of kilograms is about to be dethroned. And crucially much of the key work that has led to the toppling of the Paris kilogram has been carried out at the National Physical Laboratory where the late Bryan Kibble invented the basic concepts of the device that will replace that ingot in the Pavillon de Breteuil. The Kibble balance works by measuring the electric current that is required to produce an electromagnetic force equal to the gravitational force acting on a mass. A second stage allows the electromagnetic force to be determined in terms of a fundamental constant known as the Planck constant which will, in future, be used to define a kilogram. These machines will provide the standard for weighing objects -- and that means no more dusting of old lumps of alloy to ensure they stay pure and accurate.
[...] "One key reason for doing this work is to provide international security," says Robinson. "If the Pavillon de Breteuil burned down tomorrow and the kilogram in its vaults melted, we would have no reference left for the world's metric weights system. There would be chaos. The current definition of the kilogram is the weight of that cylinder in Paris, after all." [...] Another major motivation for the replacement of le grand K is the need to be able to carry out increasingly more and more precise measurements. "Pharmaceutical companies will soon be wanting to use ingredients that will have to be measured in terms of a few millionths or even billionths of a gram," says Prior. "We need to be prepared to weigh substances with that kind of accuracy." Suggested reading: A thread on Twitter which discusses SI units and the redefinition of the kilogram.
[...] "One key reason for doing this work is to provide international security," says Robinson. "If the Pavillon de Breteuil burned down tomorrow and the kilogram in its vaults melted, we would have no reference left for the world's metric weights system. There would be chaos. The current definition of the kilogram is the weight of that cylinder in Paris, after all." [...] Another major motivation for the replacement of le grand K is the need to be able to carry out increasingly more and more precise measurements. "Pharmaceutical companies will soon be wanting to use ingredients that will have to be measured in terms of a few millionths or even billionths of a gram," says Prior. "We need to be prepared to weigh substances with that kind of accuracy." Suggested reading: A thread on Twitter which discusses SI units and the redefinition of the kilogram.
How do they make this work as the force of gravity is not constant over the surface of the Earth? Does it only work in one place?
You use a balance, which works by comparing weights, not a scale, that works by measuring force.
How does that NOT only tell you that you have the same weight on both sides (and by proxy the same mass, under the assumption 'g' isn't changing across the relatively small dimensions of the device) ? If that's so then you still only have a way to compare a new mass to a reference mass.
Because a balance doesn't compare two masses, it compares two forces.
In this case, on one side of the balance, the force is generated using electricity,
with parameters tracing back to fundamental constants, which have defined values.
Once you get that, the explanation on how this can be used to
define a mass normal is not that complicated.
And if you tell me "well, it's a test mass on one side, and then electromagnetic force on the other", then surely the latter is then balancing against local 'g' and we're back to the variable 'g' problem.
But g in fundamental units is m/(s*s), so does not depend on the mass of your
object used to measure it, and thus is independent of the definition of the kg.
A precise measurement of g at the location of your balance can therefore be
used in defining the kg.
I get the reasoning behind looking for an alternative, but I do not see this as a sensible solution. And anyway, isn't 1 kilogram already defined at 1 L of pure H2O?
Depends what you mean by "defined".
Getting the exact temperature and pressure correct is hard. It come down to what is more reproducible with a certain degree of precision, and what effort that entails.
As measurement has become more precise, it is observable that the "exact copies" of the official kilo are drifting slightly differently from the original.
They do need to know the local gravity to measure mass. There are absolute gravimeters that measure the local value of g by dropping an object in a vacuum chamber and measuring its acceleration to very high accuracy using a laser interferometer and an atomic clock. This does not depend on the mass of the test object by General Relatively. See the Wikipedia article on Gravimeter.
The USA is technically metric. Just about all of the US units a defined as some number of metric units. Example.1 inch is defined as 25.4 mm and has been that way since 1959. https://www.nist.gov/pml/weigh... https://www.nist.gov/physical-...
> How?
Too lazy to do dimensional analysis?
*sigh*
candela is a source that emits monochromatic radiation of frequency 540e12 hertz and that has a radiant intensity in that direction of 1/683 watt per steradian.
1 Watt = 1 joule per second,
1 Joule = 1 Newton meter.
1 Newton = 1 kg * m/s^2
QED.
So, yeah, candela is a derived unit, not a fundamental unit.
The kilogram was originally created by weighing 1 L = 1 dm3 of pure water at the temperature of maximum density (about 4degC), but it turns out that this is a fiendishly difficult measurement. Water is liquid, so you need a container, and it evaporates, its density is affected strongly by temperature and weakly by atmospheric pressure, surface tension does odd things, there's such a thing as "heavy water", and so on.
It's difficult to make this measurement to better than 1 part per million. So if two laboratories (which we for simplicity assume can measure lengths and volumes perfectly) both try to derive mass from volume using water, they will only agree to 6 decimal places.
But comparing standard kilogram metal weights can be done to micrograms, which is a few parts per billion uncertainty.
So I can weigh the metal weights relative to each other to 9 decimal places, but relative to water to only 6 decimal places. It's better for everyone if we use one of the metal weights as the definition, because that will let us weigh other metal weights to 9 digits, without affecting weighings of water (which will still be accurate to 6 places).
Metrology standards are routinely redefined in this way when new technology comes along which permits measurements relative to a new standard more precisely than was possible using the old standard. Some scientists work very very hard to measure the new and old standards relative to each other to a precision greater than any previous measurement relative to the old standard, so that no previous measurement is invalidated by the change.
This has already happened to the kilogram. The water-based definition was decided on in 1795. In 1799, after having spent a few frustrating years weighing water, a platinum kilogram weight was created as the standard to be used from then on. (The "Kilogramme des Archives". Platinum was chosen because it's very dense, minimizing "air bouyancy" corrections, and because it's extremely chemically inert, so doesn't rust or corrode.) But pure platinum is a bit soft, and the "Kilogramme des Archives" was getting dinged during weighings.
So in 1875 a new kilogram (the "international prototype kilogram") was made out of a platinum-iridium alloy, which has all of platinum's advantages and is much harder to damage.
Anyway, although we can measure metal weights relative to each other to 1 part per billion, it turns out that if you take two identical such weights, store them very very carefully under identical conditions for 50 years, and then re-weigh them, the relative weights have changed by up to 50 parts per billion!
This is a big problem. We don't know what is causing that change (one plausible suggestion is carbon soot and mercury pollution in the air has been sticking to the surface of the weights) or how much any single weight has changed (we can only measure they relativechanges), but clearly at least some of the weights have changed by at least 50 ppb over the last half-century.
So that is a fundamental limit on how accurately any past measurement in kilograms has been.
The new definition is actually not as good as 1 ppb in a single day, and we'll continue to use metal weights for day-to-day operations, but has the big advantage that it doesn't change over time, so in 20 years' time we'll still be able to reproduce it to 10 ppb accuracy.
These days, we know the maximum density of water isn't quite 1 kg/L (it's 999.97495 g/L at 3.983035degC when using VSMOW). But it's equal within the accuracy of any measurement made prior to the redefinition of the kilogram in 1799, so the redefinitions hurt nothing (and helped a lot).