The kilogram is getting lighter?

Source: NY Times

[i]Scientists Struggling to Make the Kilogram Right Again

RAUNSCHWEIG, Germany — In these girth-conscious times, even weight itself has weight issues. The kilogram is getting lighter, scientists say, sowing potential confusion over a range of scientific endeavor.

The kilogram is defined by a platinum-iridium cylinder, cast in England in 1889. No one knows why it is shedding weight, at least in comparison with other reference weights, but the change has spurred an international search for a more stable definition.

“It’s certainly not helpful to have a standard that keeps changing,” says Peter Becker, a scientist at the Federal Standards Laboratory here, an institution of 1,500 scientists dedicated entirely to improving the ability to measure things precisely.

Even the apparent change of 50 micrograms in the kilogram — less than the weight of a grain of salt — is enough to distort careful scientific calculations.

Dr. Becker is leading a team of international researchers seeking to redefine the kilogram as a number of atoms of a selected element. Other scientists, including researchers at the National Institute of Standards and Technology in Washington, are developing a competing technology to define the kilogram using a complex mechanism known as the watt balance.

The final recommendation will be made by the International Committee on Weights and Measures, a body created by international treaty in 1875. The agency guards the international reference kilogram and keeps it in a heavily guarded safe in a château outside Paris. It is visited once a year, under heavy security, by the only three people to have keys to the safe. The weight change has been noted on the occasions it has been removed for measurement.

“It’s part ceremony and part obligation,” Dr. Richard Davis, head of the mass section at the research arm of the international committee.

“You’d have to amend the treaty if you didn’t do it this way.”

That ceremony has become a little sorrowful as the guest of honor appears to be, on a microscopic level at least, wasting away.

The race is already well under way to determine a new standard, although at a measured pace, since creating reliable measurements is such painstaking work.

The kilogram is the only one of the seven base units of measurement that still retain its 19th-century definition. Over the years, scientists have redefined units like the meter (first based on the earth’s circumference) and the second (conceived as a fraction of a day). The meter is now the distance light travels in one-299,792,458th of a second, and a second is the time it takes for a cesium atom to vibrate 9,192,631,770 times. Each can be measured with remarkable precision, and, equally important, can be reproduced anywhere.

The kilogram was conceived to be the mass of a liter of water, but accurately measuring a liter of water proved to be very difficult. Instead, an English goldsmith was hired to make a platinum-iridium cylinder that would be used to define the kilogram.

One reason the kilogram has lagged behind the other units is that there has been no immediate practical benefit to increasing its precision. Nonetheless, the drift in the kilogram’s weight carries over to other measurements. The volt, for example, is defined in terms of the kilogram, so a stable kilogram definition will allow the volt to be tied more closely to the base units of measure.

A total of 80 copies of the reference kilogram have been created and distributed to signatories of the metric treaty. The sometimes colorful history of these small metal cylinders underscores how long the world has used the same definition of the kilogram.

Some of the metal plugs were issued to countries that later vanished, including Serbia and the Dutch East Indies. The Japanese had to surrender theirs after World War II. Germany has acquired several weights, including the one issued to Bavaria in 1889 and the one that belonged to East Germany.

To update the kilogram, Germany is working with scientists from countries including Australia, Italy and Japan to produce a perfectly round one-kilogram silicon crystal. The idea is that by knowing exactly what atoms are in the crystal, how far apart they are and the size of the ball, the number of atoms in the ball can be calculated. That number then becomes the definition of a kilogram.

To separate the three isotopes of silicon, Dr. Becker and his team are turning to old nuclear weapons factories from the Soviet Union, where centrifuges once used to produce highly enriched uranium are able to produce the required purity of silicon.

“We need so many nines,” Dr. Becker said, and Soviet uranium processors are one of the only places to get them. “With the Russians, we’re getting about four of them,” or 99.99 percent pure silicon 28.

A test crystal has already been produced, and Dr. Arnold Nicolaus, another scientist at the German standards laboratory, is responsible for measuring whether it is perfectly round. He has measured the crystal in a half-million places to determine its shape.

It’s probably the roundest item ever made by hand. “If the earth were this round, Mount Everest would be four meters tall,” Dr. Nicolaus said. An intriguing characteristic of this smooth ball is that there is no way to tell whether it is spinning or at rest. Only if a grain of dust lands on the surface is there something for the eye to track.

Scientists from the United States, England, France and Switzerland say the challenge of calculating the precise number of atoms in a silicon crystal is too imprecise with today’s technology so they are refining a technique to calculate the kilogram using voltage.

“Measuring energy is easier than counting atoms,” said Dr. Richard Steiner, a scientist at the National Institute of Standards and Technology in Washington, who is leading the international project to create the watt scale.

In the last few weeks, he has reported that his experiments have yielded data that are close to what they need. “Now we’re into the picayune, itsy-bitsy errors,” he said, having recently corrected “totally ridiculous” errors of 100 parts per million.

The idea of the watt balance is to measure the electromagnetic force needed to balance a reference kilogram. As long as the gravitational field is precisely known for the location of the experiment, the mass on the scale can be related to power. (The gravitational field is a complicated calculation that needs among other things constantly updated changes in tidal forces.)

The definition of the kilogram would then be a measurement of that power or in terms of something that could be derived from it, like the mass of an electron. The experiment in Washington is occurring in a large three-story structure, but in spite of the complexity and circuitous route of calculating mass, Dr. Steiner says he is confident that his team will have persuasive data shortly.

“In the short term, I think we’ll win,” he said.

Dr. Davis, who is working closely with those making the final decision about the fate of the kilogram, says he is not so sure. “In terms of published results, the watt balance is closer of the two,” he said. “But it’s very hard to say which is better.”

Many scientists believe that the most elegant way to define the kilogram is by counting out a kilo’s worth of atoms of an element. A project is under way to test that with gold atoms. But the sheer number of atoms in a kilogram, a number with roughly 25 digits in it, makes that approach unfeasible for the foreseeable future.

For now, Dr. Davis is willing to set his sights lower in the error-prone world of superprecision measurements. “It would be nice,” he said, “just to have two experiments in the world that agreed with each other.”[/i]

Woh :eek:

I never thought there would be so much involved in setting a standard measurement value - all that just to measure my weight :stuck_out_tongue:

I have a simple and easy solution.
Make 1Kg = 1000.000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000g.

You can add as many 0’s as you want :slight_smile:
The ball will win, cause volts is geeky

i hope they make it a bit smaller…make me weight less


you see, this is exactly the reason why i like the english measurments better. You don’t hear about the definition of a pound changing, do you?:bigsmile:

I didn’t know there was such a thing as a definition of a english pound :bigsmile:

Who else here thinks that the platinum-iridium cylinder is losing its mass because of molecular diffusion at its surface? They should refridgerate it to lower the overall distribution of particle energies, to reduce diffusion.

Originally posted by Devils Advocate
Who else here thinks that the platinum-iridium cylinder is losing its mass because of molecular diffusion at its surface? They should refridgerate it to lower the overall distribution of particle energies, to reduce diffusion.

Yes, they could either spend lots of money on refriderating it, preferably at absolute zero. Or they could spend lots on redifining it. Wither way money goes down the tube.

imagine going to the doctor, and having him hook a positive / negative to you, and finding your weight…

//thinks the ball wins

Originally posted by kwkard

Yes, they could either spend lots of money on refriderating it, preferably at absolute zero. Or they could spend lots on redifining it. Wither way money goes down the tube. [/B]

Of course. My suggestion was meant to delay the weight loss of the 1kg standard. Until such time that we do have an acceptable standard, we will be stuck with keeping that lump of metal from losing weight :eek:

Submits another proposal for measuring a 1kg mass:

We know the equivalent energy of 1kg of mass by E = mc^2.
For a reasonable duration, T, we can also know the equivalent power of a laser, P, which would effectively shoot out the equivalent energy E if its power is kept constant over duration T. (We may want to use nanograms or very long T’s to reflect the enormity of the c^2 factor).

Since we already have a standard length scale, we then produce a scale balance which would recreate a moment produced by 1 kg an abitrary distance from the pivot. Because both the pivot distance from the photonic pressure and the (simulated) mass moment is arbitrary, we can therefore tailor this mechanism to produce the (simulated) moment, and thus (simulated) mass, of 1kg.

N.B. Reproducing the simulated 1kg moment is not very necessary, as long as you have a power (or any quantity, really) you know corresponds to 1kg, you can design any other means of reproducing the 1kg. The balance mechanism was included to illustrate how you can reproduce the mass from power.

What do you guys think?

Yeah, I’m bored :stuck_out_tongue:

Why not use a (super)computer to recalculate the mass-to-energy and energy-to-mass conversions over the period of time it is re-weighted to compensate for the loss.

Seems like the fastest and cheapest method to me.

Originally posted by Mr. Belvedere
[B]Why not use a (super)computer to recalculate the mass-to-energy and energy-to-mass conversions over the period of time it is re-weighted to compensate for the loss.

Seems like the fastest and cheapest method to me. [/B]

The mass isn’t lost via mass to energy conversions. It is lost (IMO, anyway) via molecular diffusion. Metal solid -> metal vapour -> diffusion into container walls -> diffusion through container walls -> diffusion out of container. There are also probably aspects of a very small amount of volumetric expansion because of diffusion of gas/container particles into the metal, which would result in a small bouyant effect.

All of these effects are difficult to quantify and even harder to predict.

Incidentally, we know how much and how fast the mass is losing mass, simply by recording the test readings of the mass, otherwise we would not have noticed in the first place ;). Why not just use the original readings? Because the instruments that did measure them are not perfectly identical, have their own calibration errors. If you have to reproduce a result, it is better to reproduce the experiment than to, ehm, fudge the readings.

Originally posted by Devils Advocate

The mass isn’t lost via mass to energy conversions. It is lost (IMO, anyway) via molecular diffusion. Metal solid -> metal vapour -> diffusion into container walls -> diffusion through container walls -> diffusion out of container. There are also probably aspects of a very small amount of volumetric expansion because of diffusion of gas/container particles into the metal, which would result in a small bouyant effect.

Ok ok , i forgot mass-to-mass and energy-to-energy conversions :slight_smile:

I think in the end there will be only one recalibration (such as they did with the time , date and other things) , which will be needed only once per 100 years or so.

Better make a time capsule to remind or greatgrandchildren that they need to recalibrate their measaruments.

i read some where…that the position of the world is changing (north and south pole/grafitation field position)…therefor weight will change also

What about this way? … from the National Institute of Standards and Technology

"[i]The Electronic Kilogram

Customer Needs

The kilogram is the only remaining base unit in the International System of Units (SI) whose definition is based on a physical artifact rather than on fundamental properties of nature. Environmental contamination or material loss from surface cleaning, or other unknown mechanisms, are causing the mass of the kilogram to vary by about 3 parts in 108 per century relative to sister prototypes. This observed drift highlights a significant shortcoming of the SI system. The measured values of many physical constants are based on mass, and these constants are regularly used in quantum-based measurement systems, such as the Josephson effect, which are becoming more significant to the growth of international technology and trade accreditation. Thus, with a time-drifting mass standard, adjustments to the value of physical constants must be made periodically to maintain the consistency of the SI system. Moreover, each future change will adversely affect a continuously growing technology base that relies increasingly on electronic testing, quality control, and environmental monitoring. The adoption of the electronic kilogram as the mass standard will improve the consistency of the SI and will also provide better determinations of many fundamental physical constants, such as the charge and mass of the electron, that serve the general scientific and technological communities.

Technical Strategy

The equivalence of electrical and mechanical power provides a convenient route to the measurement of mass in terms of other quantum mechanically defined measurement units. The apparatus at the Electronic Kilogram facility is a balance that compares both kinds of power in a virtual measurement that is unaffected by the dissipative forces of friction and electromagnetic heating. The experimental observables are length, time, voltage, and resistance. These quantities are all measured with respect to fundamental and invariant quantum phenomena: atomic clocks, lasers, the Josephson effect, and the quantum Hall effect.

It is necessary to reduce the total measurement uncertainty of this experimental apparatus by a factor of 10 to the level of 0.01 ppm to monitor the mass of the kilogram artifact mass standard. A substantial upgrade of the facility to reduce many known sources of error has been accomplished to achieve this goal. Experimental operations this year included extended tests at full magnetic field and combined-mode watt data acquisition in vacuum. At years end, a new induction coil was being wound. With an ultra-stiff ceramic form, this coil should reduce ground vibration effects. Also, several custom-built electronic current sources and pre-amps were being tested for use in producing a quieter, more stable magnetic field and higher resolution voltage measurement.

DELIVERABLES: By 2003, establish the performance level of the full system with respect to long-term stability of alignment, data acquisition, and reference standards and determine the uncertainty achievable for the watt realization.

DELIVERABLES: : By 2004, optimize the system for regular monitoring of the kilogram at an uncertainty level of 0.01 ppm.

Related to this work is a multi-lab effort to provide the measurement of micro-scale forces that are traceable to the International System of Units (SI). The accurate realization and measurement of micro- and nano-Newton level forces requires the development of a new kind of force comparator housed in an appropriate laboratory environment with vibration isolation, climate control, and low airborne contaminant levels. This five-year competence project, the Microforce Realization and Measurements Project (MFMP), includes researchers from the Automated Production Technology and Precision Engineering Divisions of the Manufacturing Engineering Laboratory (MEL), from the Ceramics Division of the Material Science and Engineering Laboratory (MSEL), and from the Electricity Division.

DELIVERABLES: : By 2003, complete the development of the preliminary electro-mechanical force balance and demonstrate the comparison of mechanical to electrical forces at a level of 10-5 N with a total relative uncertainty of 10-3.

DELIVERABLES: : By 2005, establish a metrological basis for small force measurement by developing an electronic realization of force traceable to the SI in the regime between 10-8 N and 10-4 N.


The watt balance system became fully operational using both volt/velocity and force/current modes. A new induction coil was installed and tested in air for about four months of shakedown operation at full magnetic field. This second induction coil was redesigned to be stiffer and equipped with the interferometer mirrors in optimal positions to reduce the effect of internal vibrations producing uncorrelated voltage pickup. Results indicated an improvement of 10 times less noise at vibration frequencies near 20 Hz. However, reinforcing fiber, epoxy, and a conductive coating meant to help in the coil’s stiffness and anti-static buildup inadvertently caused severe leakage resistance. A new coil form composed of interlocking ceramic pieces was designed to address these problems and further increase the stiffness. It will replace the existing coil by December, 2002.

The in-air testing was followed by six weeks of testing in vacuum. All reference values were available for data analysis, including a programmable Josephson array and a GPS receiver. The measurements demonstrate that the average value of the daily data runs was very stable; the standard deviation of the runs generally ranged from 0.2 ppm to 0.4 ppm, uncorrected for tides. While not yet an improvement over the measurements routinely achieved leading up to the 1998 results, many control parameters have not yet been optimized. Because there are known sources of error at present (e.g., serious leakage resistance in pickup-coil, small instabilities of the magnetic field, and some remaining issues with the data analysis algorithms) it is too early to make any comparison of the present measurements with the previously reported realization of the watt. Some new electronics for reducing the noise in the magnetic field control are now ready for on-line testing, and recent programming optimization has reduced the signal/noise ratio in the volt/velocity ratio measurement.

The upper curves show the voltage (o) and velocity (´) during a single sweep (~75 seconds duration) of the moving coil through the radial magnetic field. The relatively large scatter results primarily from relative differential vibration between the coil and the field. The lower curve shows the volt/velocity ratio, which should be nominally constant. Because the vibrations are common to both volt and velocity measurements, the scatter in the ratio is reduced by about 2 orders of magnitude. The ratio shows the typical “M”-shaped profile of the magnetic field.

Some long term testing has given reassurance that the alignments of the superconductor, coil, and laser optics are stable and reproducible. This was accomplished in part by improving the laser detection scheme for alignment positioning and also the room temperature control. By inserting the alignment system’s diode laser beam into fiber optic cable, the number of guiding elements was reduced to one rigidly held launch fixture directly at the beginning of the reflector path. This eliminated the physical drift of multi-element mirror guides. Also, the building temperature controller was rebuilt and reprogrammed, now maintaining excellent thermal stability inside the balance room. This computerized measurement and feedback system overrides measurement and feedback of the pre-existing thermo-couple/pneumatic system to obtain long-term temperature control well within 0.2 C.

For the Microforce Realization and Measurements Project, a preliminary microforce balance calibration system was completed this year, with results reported in a paper submitted at the Conference for Precision Electromagnetic Measurements 2002 in Ottawa Canada. Measurement tests at 10 µN, 100 µN, and 200 µN resulted in discrepancies between mechanical and electrical components at parts in 104 with uncertainties of similar value. This system was used to calibrate a commercial piezo-resistive force cantilever to an uncertainty of about 2 %. This result is several times better than the 10 % specification of the instrument.[/i]

Actually, Mr. Belvedere, you reminded me about something. Radioactivity will actually result in a loss in mass of the lump o’ metal too. If you get fission of the gold/platinum because of it being hit or energised by a radioactive source (however remote) or background radiation, part of the binding mass will be lost. Mass in the form of metals being converted to lighter gases will also be lost.

Interesting how things become so much more complicated when the nth zero counts.:eek:

i knew that…

Or maybe it would be possible to put some amount something fissile material on nuclear reactor where it will lose some of its mass due to e=mc^2. You could first measure the fissile materials mass before putting it in reactor and then measure the energy that is produced (its not 100% correct word but i dont remember better) and finally the mass of the fissile material after the process. Then you would calculate the energy required to form 1kg of matter. :bigsmile:

Oh my god this gives me a headache.
If all the units of measurement are interlinked, volt, kilogram, litre, how the feck are you supposed to find somewhere to start?

And is it just me being stupid, or - should they not attempt to define the gram rather than the kilogram? This way building it out of a certain number of atoms would be 1000x easier.:stuck_out_tongue:

It’s only a matter of time before they whinge about atoms being different weights, so they have to add up single quarks or something.
Perhaps they could have a neurotic woman stand on some scales four times a day, and assess a kilogram in relation to litres of ice cream input.