# The Kilogram

Alfi
Just stumbled on this in Wiki and wondered.
The kilogram or kilogramme (symbol: kg) is the base unit of mass in the International System of Units (known also by its French-language initials “SI”). The kilogram is defined as being equal to the mass of the International Prototype Kilogram (IPK; known also by its French-language name Le Grand K), which is almost exactly equal to the mass of one liter of water.[1] It is the only SI base unit with an SI prefix as part of its name. It is also the only SI unit that is still defined in relation to an artifact rather than to a fundamental physical property that can be reproduced in different laboratories.

It is also the only SI unit that is still defined in relation to an artifact rather than to a fundamental physical property
Is there a good reason why. A lot of measurements are based on this artifact.

Certainly a lot of diets are. :)

I found this bit further down the page interesting.

Ultimately, the watt balance would define the kilogram in terms of the Planck constant, which is a measure that relates the energy of photons to their frequency. The Planck constant would be fixed, where h = 6.62606896 × 10–34 J·s (from the 2006 CODATA value of 6.62606896(33) × 10–34 J·s) and the kilogram would be defined as “the mass of a body at rest whose equivalent energy equals the energy of photons whose frequencies sum to 1.356392733 × 1050 Hz.”[29]

Last edited by a moderator:

Redbelly98
Staff Emeritus
Homework Helper
Standards are based on the most accurate measurement technology available. That is why, for example, the definition of the meter was changed from a wavelength standard to a definition based on the speed of light and the second. Technology is such that time can be measured much more accurately than distance, hence the redefinition of the meter in terms of time.

The fact that the kg is based on a prototype reflects the difficulty in using some fundamental standard, such as 1 mole of Carbon-12 atoms. How do we count 1 mole of C-12 atoms?

By the way, efforts are underway to redefine the kg in terms of a silicon atoms.
http://www.iop.org/EJ/abstract/0026-1394/40/6/008

f95toli
Gold Member
The definition of the the kilogram won't change any time soon. The definitions used in the SI can only be changed at a conference which is held in Paris every 5-10 years or so (I don't think there is a fixed interval); the next meeting is pretty soon (2009?) and since nothing will happen then it will take at least another ten years or so before anything changes (there are, however, a few definitions that WILL be changed at the next meeting).

An Si artifact might be ultimately be better than the current artifact but most metrologists are not happy about just changing from one artifact to another.
Ultimately we would like to have a "natural" standard but so far no one has come up with a method that works well enough. There are two Watt balances in the world (one in the US and one here in the UK) but the latest (preliminary) values from them unfortunately do not agree (and the disagreement is actually larger than the error bars).
Moreover, the people working on them are finding it increasingly hard to get funding; meaning it is very unlikely that the work will result in a new standard.

There is also a lot of work on new methods for cleaning the artifact in Paris; most of the drift in the kg is likely to be from dirt accumulating on its surface.

What was undesirable about the original definition of the gram and kilogram as the mass of a certain volume of water at STP? Doesn't the precision of the meter now make volumes also precise?

The definition of the the kilogram won't change any time soon.

well, i certainly believe i will live to see a change in definition away from the BIPM prototype (since the mass of that prototype, relative to the masses of copies they have made, is varying by 50 ppb, which is about the present accuracy of measuring Planck's constant, and, i think accounts for it) and i'm 52 years old. we'll see a change in the coming decade or two. i think NIST heavyweights like Peter Mohr and Barry Taylor are agitating for a change. could happen sooner than you think.

An Si artifact might be ultimately be better than the current artifact but most metrologists are not happy about just changing from one artifact to another.

of course! that's why they need to change the definition away from an artifact and to a readily duplicated physical experiment.

Ultimately we would like to have a "natural" standard but so far no one has come up with a method that works well enough. There are two Watt balances in the world (one in the US and one here in the UK) but the latest (preliminary) values from them unfortunately do not agree (and the disagreement is actually larger than the error bars).

so which one is the one used for the accepted value of Planck's constant (in terms of the current kg)? the current value of h is good for better than 50 ppb (in terms of the present kg), which is, i think, due to the variation of the kg artifact (relative to the copies).

Moreover, the people working on them are finding it increasingly hard to get funding; meaning it is very unlikely that the work will result in a new standard.

There is also a lot of work on new methods for cleaning the artifact in Paris; most of the drift in the kg is likely to be from dirt accumulating on its surface.

they have not established that. they have a cleaning regimin that cleans those artifacts pretty good (and are believed to not strip off Pt-Ir atoms) and the std-deviation is still around). i think that, if they can make repeatable copies of the watt-balance which they would have to get a decent consistent measure of Planck's constant, they could use the same instruments to reverse the roles: define Planck's constant to be whatever best measured value they have at the time of changed kg definition, and then define the kg to be whatever mass it has to be to make Planck's constant to be that defined value, along with c = 299792458 m/s and the second to be 9192631770 133Cs "hyperfine radiation" cycles. given those three constraints, they can define the second, meter, and kilogram. but the problem is technological. the watt-balance does not, at present, make for an economical and consistent "transfer standard", whereas kg prototypes do (but they're not so consistent after all).

BTW about the meter, a useful reference is
Time Line for the Definition of the Meter

by switching from a "wavelength" standard to a "speed of light" standard, what they really did was switch from a 86Kr standard for length and a 133Cs standard for time, to using the same 133Cs for both length and time. since the speed of propagation of radiation for 133Cs and 86Kr is the same, the fact that there was measurement variance in the measure of c before 1983 really simply reflects a measurement error in the relative wavelengths or frequencies of 133Cs and 86Kr (which are 9.19263177x109 Hz and 4.94886516194x1014 Hz respectively). after 1960, when they first moved the meter definition away from the BIPM meter prototype, any variance in the measure of the speed of light really meant a variation in the relative measure of those two frequencies.

f95toli
Gold Member
well, i certainly believe i will live to see a change in definition away from the BIPM prototype (since the mass of that prototype, relative to the masses of copies they have made, is varying by 50 ppb, which is about the present accuracy of measuring Planck's constant, and, i think accounts for it) and i'm 52 years old. we'll see a change in the coming decade or two. i think NIST heavyweights like Peter Mohr and Barry Taylor are agitating for a change. could happen sooner than you think.

Maybe. But the problem is that there is no way NIST can gain enough support unless the experiments have been duplicated by at least one other NMI. Moreover, as far as I can tell the politics is quite complicated (although I only know what I hear in our tea room) so it is difficult to tell.

of course! that's why they need to change the definition away from an artifact and to a readily duplicated physical experiment.

I agree. But which kind of experiment? The problem with a Watt balance is that it is extremely complicated (I have seen the setup), and it requires a great deal of expertise from many different people to run which in turn means that errors are difficult to track down.
Moreover, it is expensive and -as I wrote above- it seems to be hard to get funding. The Watt balance in the UK (NPL) was actually due to close down a few months ago, the project has been extended by a few months but once they have a new (and hopefully better) value they will have to close down permanently; there is no money for a new generation (and NPL is leaving the building where the current balance is located, there are only a couple of experiments left there).

so which one is the one used for the accepted value of Planck's constant (in terms of the current kg)? the current value of h is good for better than 50 ppb (in terms of the present kg), which is, i think, due to the variation of the kg artifact (relative to the copies).
As far as I know, neither. The accepted value is a CODATA value, i.e. it is not determined by a single type of measurement (although I believe Watt balance data has been used). As far as I know the latest Watt balance values (the ones that were published in Metrologia a while ago) are not used due to the fact that the NIST and NPL values do not agree; i.e there is a serious problem with one of the experiments and we still don't know which one.

they have not established that. they have a cleaning regimin that cleans those artifacts pretty good (and are believed to not strip off Pt-Ir atoms) and the std-deviation is still around).

Some people think that improved cleaning might help. They recently started trials where they are using ultrasound to clean it (although of course they are using one of the copies for testing). But again, I am not an expert in mass (or lenght) metrology so this is just what I have been told/read.

Gold Member
The definitions used in the SI can only be changed at a conference which is held in Paris every 5-10 years or so (I don't think there is a fixed interval); the next meeting is pretty soon (2009?) and since nothing will happen then it will take at least another ten years or so before anything changes (there are, however, a few definitions that WILL be changed at the next meeting).

What is likly to change. Its just annoying stuff like that happning when your learning physics (I should be at unie by then).

f95toli
Gold Member
Nothing that will have any practical implications unless you happen to work at an NMI (although most people like to idea of a neater SI).
As far as I understand the Kelvin (by fixing Boltzmanns constant) and the Mole (by fixing Avogadros constant) are on track to be re-defined.

Steps have also been taken to re-define the ampere in terms of the elementary charge; but so far no experiment (electron pumps,counters etc) is sensitive enough, so that will have to wait, at least until the next conference (but it will happen eventually)

Redbelly98
Staff Emeritus