Redefining the Kilogram | Hacker News Reader

Redefining the Kilogram

scientificamerican.com

61 pointsbushido46 comments

There's also an excellent video from Veritasium, a science and education YouTube channel, about what's being done to define the kilogram

https://www.youtube.com/watch?v=Oo0jm1PPRuo

Also interesting 'Practical Engineering' building a DIY watt scale: https://www.youtube.com/watch?v=ewQkE8t0xgQ

For anyone interested there's an excellent BBC4 documentary on this very subject by Professor Marcus du Sautoy. It's called Precision: The Measure of All Things, unfortunately isn't on iPlayer at the moment but it can be found elsewhere, including DailyMotion.

https://www.dailymotion.com/video/x2hy5l5

It really is shockingly good. Despite the slightly mundane title, it's interesting and understandable all the way throught. Highly recommended.

For those actually curious of the details of all those changes, Wikipedia has a good overview:

https://en.wikipedia.org/wiki/Proposed_redefinition_of_SI_ba...

How were these working copies made?

The article makes it seem like there's some "copy" operation you can apply to the reference artifact, but that seems absurd...

I Googled "working standard kilogram" (no quotes), in case I'd find some article, and I did find only this:

https://www.technology.matthey.com/article/17/2/66-68/

I don't really get a sense after reading, of how these things are used to make more of them.

I would assume you just make an object that will be the working copy, and align it's weight with that of the standard.

How do they align its weight?

Same way as you machine anything. Start with more material than you need and remove until it is the correct dimensions. In this case, start slightly heavier and polish until it is the correct weight.

The first google term you don't have is "traceable mass calibration NIST" that will lead you to the NIST price list and terms and conditions.

They do "double substitution procedure" (another google term) for most masses for about three grand plus or minus incidentials (cleaning, shipping). Unfortunately the best descriptions I can find of NIST double substitution procedure is docx unfree format.

However it boils down to a really high quality double arm balance (like a lever with a pivot in middle and buckets on each end) and its temp controlled to keep the lengths of the arms identical and they stick your (test) mass on one side, stack calibration weights on the other side and record what balances, then remove your mass and stack more calibration masses in its place optimistically very similar in mass to the other side, but it won't be equal because one arm is probably a nanometer longer than the other arm or something. Essentially they're testing it twice in order to factor out a difference in arm lengths however small. You can get around eight or nine sig figs of accuracy now a days using this procedure.

I would enjoy seeing an electrostatic force balance like they use for microgram scale masses.

It's hard to imagine that a physical lever-based mechanical scale is so accurate, never would have imagined. Thanks for the detailed comment.

NIST has also just released its own article about the redefinition: https://www.nist.gov/si-redefinition/kilogram-kibble-balance

It has quite a few details that give a sense of the engineering challenges involved.

- The Kibble balance has a weighing mode, and a calibration mode; the two are used to cancel out a significant portion of the systematic error introduced by the mechanical system

- Measurements are made in a vacuum to eliminate effects of air buoyancy, which are significant at the desired level of precision.

- One advantage of the wheel configuration is that it allows the supporting cables to move the coil exactly vertically, thereby avoiding the undesirable sideways motion that a coil suspended from a beam balance can experience.

I could go on, but this gives an idea and you can go read it yourself ;)

I think this is good. Not a physicist but a lot of the SI definitions seem contrived and arbitrary [1] rather than based on physical fundamentals.

[1] https://en.wikipedia.org/wiki/SI_base_unit

This will still have some arbitrary constant, much like the meter and the second.

Indeed, it's not the arbitrariness that's going away, but the need for special artifacts such as the kg standard.

One can be even more specific: the need for the last remaining special artefact that is the kg standard.

Fun fact: the metre was was originally defined to be 1/10 millionth the distance from the pole to the equator (so that the circumference of the Earth would be 40 million metres).

Sadly that list is based on data from the 2005 standard. Several of these have since been redefined.

Can you point to one in particular? We can update the article if it needs to.

I found this: https://www.bipm.org/en/publications/si-brochure/

Also includes the (currently in draft) 9th edition.

The definition of the second has changed between the 8th edition of the SI Brochure last updated in 2014, with which Wikipedia agrees:

> 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 caesium 133 atom.

and the draft 9th edition:

> The second, symbol s, is the SI unit of time. It is defined by taking the fixed numerical value of the caesium frequency ∆νCs, the unperturbed ground-state hyperfine transition frequency of the caesium 133 atom, to be 9 192 631 770 when expressed in the unit Hz, which is equal to s−1 .

The draft 9th edition incorporates the work to redefine several of the base units in terms of fundamental physical constants and dispenses with the last remaining reference artefact. The changes come up for a vote next month.

* https://www.bipm.org/en/measurement-units/rev-si/#when

So Wikipedia could include the concrete wording of what the redefinitions are currently intended to be. And it does.

* https://en.wikipedia.org/wiki/Proposed_redefinition_of_SI_ba...

It's inline with the 8th edition of the SI Brochure, published in 2014. It's not inline with the draft 9th edition.

It will be always be arbitrary anyway. Say we used the wavelength of some special light as unit for length... Why did we choose that light, and not something else?

The point with the kilogram redifinition is that it's currently tied to some piece of metal stored in Paris somewhere. We can physically lose the definition.

And we actually do:

„Though the IPK, the current primary artefact, and its replicas are stored in carefully controlled laboratory conditions, their masses have been subject to fluctuation as a result of poorly understood factors, possibly including handling, cleaning and contamination. The IPK has diverged from its replicas by 50 μg since their manufacture late in the 19th century.“

https://en.m.wikipedia.org/wiki/Kilogram

We do use resonant frequency of certain atoms to define units, but it is not arbitrary, rather it was a transition chosen for it's extremely small width and ease of measurement. It is quite a bit less arbitrary than using the size of our planet or of an artifact we have made.

And we are getting even better, with definitions based on fundamental constants now, independent of a particular measurement experiment. This is as non-arbitrary as it can get (we will continue using a conversion factor to make the unit human scaled, but that is all).

But one meter is still the length we are used to.

So we got the resonant frequency of some atoms, which is arbitrary because they just happened to be relatively easy to measure for us... Multiplied by some arbitrary factor so the unit reaches the usual length.

That doesn't seem very fundamental to me. Could you imagine proposing that to some intergalactical economic union as new standard?

The resonant frequency of the atom is not arbitrary, it is a physical constant.

That's the key and the point: It allows to measure exactly 1 second anywhere in the universe because the resonant frequency of that atom is the same anywhere in the universe.

Compare with a previous definition of the second as a fraction of the solar day... A solar day isn't in fact constant and, in any case, once you leave earth it means nothing at all.

The definitions are still arbitrary. The important change is that they are arbitrary numbers of a fundamental physical constant instead of being completely arbitrary.

That means that the unit may be exactly measured anywhere in the universe based on the definition only without the need of any artifact.

Yeah, but there are interesting criteria: an abstract definition for theoretical re-construction of reference artifacts(I don’t know if anyone has ever actually re-built a reference without comparing it to a previously existing reference during development), using integers in the definition where it makes sense, etc.

They were created well before the natural units or atomic units were discovered so they are going to based off of arbitrary measurements that were repeatable by scientists of the past. SI is just as arbitrary as imperial but at least it has the advantage of base-10 for the prefixes. The problem is that you can't just change the kilogram to whatever nice natural unit you want overnight so even moving to nice universal constants would still give you some weird coefficients.

>it is roughly five parts in 100 million heavier than all the working standards, which have been leaving behind a few atoms of metal every time they are put on scales

Are the working copies really only shedding such a small number of atoms per use? Just seems like it should be a higher number.

Platinum has molar mass ~195g, Iridium ~192g. This means the working copies have between ~5.128 and ~5.208 mol atoms in a kg. Let's round it to 5.2 mol, that's ~3.13 * 10^24 atoms.

Five parts in 100 million is ~1.565 * 10^18 atoms. Sound like a pretty large number to me.

Well some back of the napkin math, atomic mass of platinum, the heavier constituent of the standard is 195. So more than 5 mols of these atoms are in the block, because 1 mol of pt is 195 grams. So there’s around 5610^23 atoms. So 5 parts per 100 million adds up to a lot. Maybe 1.5*10^15 atoms? It’s been a while since I had chemistry...

Can somebody explain to me why we need both kilogram and mole (https://en.wikipedia.org/wiki/Mole_(unit))? One can be defined in terms of the other.

Kilogram of any substance weighs the same, whereas the moles of different substances have different weights (numerically equal to the molecular weight of the substance).

Sure, but can't the kilogram then be easily defined in e.g. carbon-moles?

it could (again, based on a particular isotope), but only in theory, because how would one count the mole worth of atoms?

isn't one mole of carbon-12 by definition 12 grams?

Unit conversion no longer being an issue, I wish everyone went back to the imperial units. (Computers are not 10-based anyway.)

I'm trying to imagine what the benefit would be. Could you explain?

Parties of two, three, four, eight, and twelve kids can now look forward to simple divisions of cake.

You can "spend a penny" again. Although I haven't seen a pay toilet in over 30 years. (I don't mean the ones you put money on a plate at the entry, I mean one you put money INTO the door)

What do pennies (or pence) have to do with imperial units? A penny is exactly 1/100 of one pound sterling or GBP 0.01, which is a nice metric style power of ten division of the larger unit. If you want monetary units more congruent with imperial, then you may be thinking of l.s.d. or librae, solidae and denarius - pounds, shillings and pence - with twelve pennies to the shilling. Note this is UK only, I don't think the USA ever had non decimal currency?

E.g., the largest economy in the world uses them.

What does the economy have to do with units of measurement? Not to mention that all of the scientific and engineering world (yes, even in the US) uses metric. If you were to do a headcount, the US is very strongly in the minority on this one. Not even the UK uses imperial outside of the historical usage of miles.

Most anti-metric folks I've spoken to don't like metric because an inch is an ugly amount of centimetres (or similar arguments). But by the same token, a centimetre is an ugly amount of inches...

(Also, most imperial units are now defined in terms of SI units.)

"Centimeter" is an long and ugly made-up word; not to say anything about all these "tera" and "nano" (not sure what the Greeks think about them).

The main point is that with the advent of automatic computers the original reason for "metrification" is no longer valid.

I'm confused, isn't "inch" also a made up word? Aren't all words made up?

> the original reason for "metrification" is no longer valid.

This seems like a silly argument given that humans still have to write these systems, and all of the engineering and physics papers use metric. So any system interacting with anything physical (which is all the systems that would care about the use of metric/imperial units) would be more easily written using metric units.

Not to mention that humans still do many tasks that computers don't, and in those cases metric still has the same upsides it has always had.

Reply on news.ycombinator.com