The Changing Definition Of 'Kilogram'

DrLudicrous writes "The NYTimes is reporting that the platinum-iridium standard mass for the kilogram is shedding at an appreciable rate -- at least compared to other reference masses. The Pt-Ir cylinder is kept in France, and measured annually, and the slight discrepancy is important because the kg is an SI base unit- thus other quantities such as the Volt are based on it. A new standard is being sought- the two frontrunners are counting the number of atoms in a perfectly spherical single crystal of silicon, and another technique uses a device known as the Watt balance."

Comments

Counting Si

[brokenbeaker]

The problem with the single crystal of silicon method, a few years ago, was that there were all these lattice vacany defects cropping up. The formation of such point vacancies is so entropically favoured that I don't think they can ever eliminate them...

How do they measure it?

[LX.onesizebigger]

My question is, how do they measure it? Using a non-decaying meter stick? How do you measure the definition of a measure?

Best units of measure

[Alien54]

I sort of like the idea of a universal unit of measure.

One nominee that is amusing is to have the basic unit of distance based on the speed of light.

One light nanosecond = roughly 11.1 inches, kinda close to a foot.

I remember how Grace Hooper used to pass out wires that were that long, just to make the point.

Any other nominees?

Why not use diamond?

[Jeremy Erwin]

Is there any physical reason (other than that small matter of cost ) that crafting a new kilogram (or more likely, gram) out of diamond would not be an ideal solution?

BTW, theNational Physical Institute has a FAQ on its Pl-Ir standard kilo.

Re:Kilogram?

[Enraged_jawa]

Hah, yeah, true, probably due to Autodesk/Autocad defaults. When I need a metric equavalent, I just measure it with my digital calipers in english, then hit the in/mm button to see the metric equavalent.Metric is much easier to work with, though.

Re:Why not...

[wass]

The answer is another question : How would you define a standard energy?

c is a constant, of course. In fact, it's used to define the meter as how far light travels in a vacuum in 1/(299,792,458) of a second. Second is defined as the time for a certain number of vibrations of a Cesium atom to occur. As per your question of relating mass to Joules, note that high-energy physicists do this all the time. They usually refer to masses of particles as MeV/c^2. And they usually work in units where c=hbar=1, thereby making distance, time, and energy all essentially the same units (easier to do calculatins that way).

One thought that jumps to mind for a standard energy interval is the lyman alpha energy width (the jump of the electron in a hydrogen atom from n=2 to n=1 where n is the energy quantum number). Or, for mass, use a standard mass of a well-defined particle such as an electron. In fact, I'm surprised that NIST doesn't do this. It might be that isolating electrons for mass measurements are too difficult (gravity is weak), but electron mass does show up in many other calculations (specific heat of degenerate electron gases, for instance). Or isolating ultra-pure hydrogen gas and spectroscopically measuring Lyman alpha is more difficult than it seems. I guess NIST wants [relatively] easy methods for measuring these quantities.

Okay, I just found this site which answers the question. They quote

This one physical standard is still used because scientists can weigh objects very accurately. Weight standards in other countries can be adjusted to the Paris standard kilogram with an accuracy of one part per hundred million. So far, no one has figured out how to define the kilogram in any other way that can be reproduced with better accuracy than this. The 21st General Conference on Weights and Measures, meeting in October 1999, passed a resolution calling on national standards laboratories to press forward with research to "link the fundamental unit of mass to fundamental or atomic constants with a view to a future redefinition of the kilogram." The next General Conference, in 2003, will surely return to this issue.

It all boils down to ability to measure the standard units to the highest precision possible. I'm actually stunned that the mass of that bar can be weighed to that precision.

As a side note, if you can come up with a better way of measuring fundamental constants, you might win a Nobel Prize. The guys that discovered the integer quantum hall effect initially published their results as a better way to measure some of the fundamental constants.

Pipe-sizes are not that simple...

[Ashtead]

Threaded pipe dimensions in inches were based on the internal diameter at some point, just to make things even more interesting. Thus, a so-called 1/2" pipe actually has an external diameter of approximately 20 mm, which translates to somewhat more than 3/4". All the common Pipe Thread sizes are this way. There has been some attempt at metricising these, at least within Scandinavia, instead of referring to terms like 1/2" or 3/4" pipe threads, terms like R15 and R20 have been seen instead. That seems to be neither here nor there.

Just to confuse the matter more, in the 1970s, it was common to use metric sizes of threaded copper pipe, which had external diameters in sizes approximating common fractions of inches: 13mm = 1/2", 16mm = 5/8" and 19mm = 3/4" just to mention some of them. These appearently were all threaded with 1mm pitch threads.

Later, these were replaced by true metric pipe sizes with compression fittings or capillary solder fittings. Now the sizes changed again, common ones are 8, 10, 12, 15, 18, 22, and 28 mm. And of course, one needed compression fittings made for 16mm and 19mm also, so as to fit the older pipes...

That's Europe. What I have seen in the US are the commonly found so-called 1/2" copper pipes with solder fittings, this is about 16mm (5/8") in diameter, so I guess they are still using internal diameter measurements. Similarly, the so-called 3/4" pipes appear to have about 21mm outside diameter.

I guess the easiest way to turn these into metric sizes would be to redefine them as 16mm and 21mm and leave it at that. At least the traditional inch-units pipe thread sizes are roughly the same everywhere!

Re:Kilogram?

[sphealey]

Then maybe america should move out of the dark ages sometime.

Funny how when the topic is software or food supplies, everyone jumps in with comments about the dangers of monoculture and the value of diversity in supply, but when the topic is the metric system there can be no deviation from the ONE TRUE FAITH.

Personally, having gone through school at a time when the US was considering a change, and having spent some time in Europe, I have no problem with the metric system. It is more convenient from some tasks, particularly in the chem lab.

But there is nothing inherently superior about a measurement system based on powers of 10. For many tasks, such as woodworking, metric measurements are far more difficult to work with than inches and 1/16th. In fact I would argue that the most "natural" base for a measurement system is 12 as it is evenly divisible by 2, 3, and 4; whereras base 10 is only divisible by 2 and 5. Thirds and fourths are very common divisions of stuff; fifths are not, so a base 12 system is more user-friendly.

That's my 0.02 euro anyway.

sPh

Re:Kilogram?

[radish]

I do get it. This already happens in the UK, it's not a problem at all. We have lots of houses which are older than the metric system (and the USA for that matter). They use imperial stuff. We have lots of new houses - they use metric. And yet I can still call a plumber and he can figure out how to fix my pipe, and my electrician is able to fix a light. Amazing.

If there was any will to do it you'd do it, which indicates there's no will. Which is fine, I don't give a toss what you measure your wooden houses in, but don't come over all "it's too haaaaaard" - you sound like a whinging kid.